diff --git a/.gitattributes b/.gitattributes index 405fdd28daadc7b9c4014fb3709e3d3a6d60b2ec..5a5f2522fa35c89eaa3113963fd4ad8d0d600c76 100644 --- a/.gitattributes +++ b/.gitattributes @@ -59,3 +59,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.mp4 filter=lfs diff=lfs merge=lfs -text *.webm filter=lfs diff=lfs merge=lfs -text data/master_gene_table_v1_1_enriched.csv filter=lfs diff=lfs merge=lfs -text +data/master_gene_table.csv filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md index 11abdb82b497053e92e2701de5eb048373d28e71..7e7348b3b7335d1ce5dd24f474d97b4af33a6b58 100644 --- a/README.md +++ b/README.md @@ -21,8 +21,8 @@ This repository is private staging. Do not make it public until redistribution p - `data/manifest.csv` - `data/manifest.parquet` -- `data/master_gene_table_v1_1_enriched.csv` -- `data/master_gene_table_v1_1_enriched.parquet` +- `data/master_gene_table.csv` +- `data/master_gene_table.parquet` - `data/embeddings//_embeddings.npz` - `data/embeddings//_genes.tsv` - `data/embeddings//_metadata.json` diff --git a/data/embedding_id_rename_map.csv b/data/embedding_id_rename_map.csv new file mode 100644 index 0000000000000000000000000000000000000000..dc348350e41d10a6b3be724a992741ac91f26177 --- /dev/null +++ b/data/embedding_id_rename_map.csv @@ -0,0 +1,69 @@ +former_processing_embedding_id,embedding_id +alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1,alphagenome_tss1mb_regulatory_tracksummary_pca256 +basenji2_tss_131kb_human_central3_v1_1,basenji2_tss_131kb_human_central3 +biobert_gene_protein_summary_v1_1,biobert_gene_protein_summary +bioconceptvec_cbow_v1_1,bioconceptvec_cbow +bioconceptvec_fasttext_v1_1,bioconceptvec_fasttext +bioconceptvec_glove_v1_1,bioconceptvec_glove +bioconceptvec_skipgram_v1_1,bioconceptvec_skipgram +cmap_gse106127_cgs_mean_landmark978_zscore_v1_1,cmap_gse106127_cgs_mean_landmark978_zscore +deepnf_string_v12_ppmi_svd_fusion_v1_1,deepnf_string_v12_ppmi_svd_fusion +deepsea_tss_1kb_epigenomic_probs_v1_1,deepsea_tss_1kb_epigenomic_probs +depmap_crispr_gene_effect_pca256_v1_1,depmap_crispr_gene_effect_pca256 +dnabert2_117m_cds_v1_1,dnabert2_117m_cds +enformer_tss_393kb_human_central3_v1_1,enformer_tss_393kb_human_central3 +esm2_v1_1,esm2 +frogs_archs4_256_v1_1,frogs_archs4_256 +gene2vec_v1_1,gene2vec +genecompass_base_gene_token_v1_1,genecompass_base_gene_token +geneformer_v2_316m_v1_1,geneformer_v2_316m +genept_ada_v1_1_conservative,genept_ada_conservative +genept_model3_gene_protein_v1_1_conservative,genept_model3_gene_protein_conservative +gtex_tissue_median_log1p_zscore_v1_1,gtex_tissue_median_log1p_zscore +helix_mrna_full_transcript_v1_1,helix_mrna_full_transcript +hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1,hetionet_nonppi_biomedical_context_tfidf_svd256 +hpa_normal_ihc_tissue_level_zscore_v1_1,hpa_normal_ihc_tissue_level_zscore +hpa_subcellular_location_multihot_zscore_v1_1,hpa_subcellular_location_multihot_zscore +hyenadna_medium_160k_cds_v1_1,hyenadna_medium_160k_cds +mahi_all_contexts_mean_v1_1,mahi_all_contexts_mean +mahi_global_v1_1,mahi_global +mashup_string_v1_1,mashup_string +msa_transformer_depth1_uniprot_human_v1_1,msa_transformer_depth1_uniprot_human +newt_archs4_256_v1_1,newt_archs4_256 +newt_cellnet_v1_1,newt_cellnet +newt_go_256_v1_1,newt_go_256 +newt_go_graph_v1_1,newt_go_graph +newt_msigdb_bundle_v1_1,newt_msigdb_bundle +node2vec_consensus_ppi_v1_1,node2vec_consensus_ppi +nt_v2_500m_multispecies_cds_v1_1,nt_v2_500m_multispecies_cds +opa2vec_goa_human_v1_1,opa2vec_goa_human +orthrus_base_4track_full_transcript_v1_1,orthrus_base_4track_full_transcript +primekg_nonppi_biomedical_context_tfidf_svd256_v1_1,primekg_nonppi_biomedical_context_tfidf_svd256 +probe_aac_v1_1,probe_aac +probe_albert_v1_1,probe_albert +probe_apaac_v1_1,probe_apaac +probe_bert_bfd_v1_1,probe_bert_bfd +probe_bert_pfam_v1_1,probe_bert_pfam +probe_blast_v1_1,probe_blast +probe_cpc_prot_v1_1,probe_cpc_prot +probe_esmb1_v1_1,probe_esmb1 +probe_gene2vec_uniprot_v1_1,probe_gene2vec_uniprot +probe_hmmer_v1_1,probe_hmmer +probe_ksep_v1_1,probe_ksep +probe_learned_vec_v1_1,probe_learned_vec +probe_mut2vec_v1_1,probe_mut2vec +probe_pfam_v1_1,probe_pfam +probe_protvec_v1_1,probe_protvec +probe_seqvec_v1_1,probe_seqvec +probe_t5_v1_1,probe_t5 +probe_tcga_embedding_v1_1,probe_tcga_embedding +probe_unirep_v1_1,probe_unirep +probe_xlnet_v1_1,probe_xlnet +pubmedbert_gene_protein_summary_v1_1,pubmedbert_gene_protein_summary +rnafm_full_transcript_v1_1,rnafm_full_transcript +scfoundation_gene_pos_v1_1,scfoundation_gene_pos +scgpt_pancancer_v1_1,scgpt_pancancer +scgpt_whole_human_v1_1,scgpt_whole_human +scprint_medium_v1_5_gene_v1_1,scprint_medium_v1_5_gene +spliceai_canonical_splice_site_profile_v1_1,spliceai_canonical_splice_site_profile +uce_33l_v1_1,uce_33l diff --git a/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/README.md b/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/README.md new file mode 100644 index 0000000000000000000000000000000000000000..be5a6a7431b61be10eec1b9428fd9dcc9cb350c8 --- /dev/null +++ b/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/README.md @@ -0,0 +1,74 @@ +# AlphaGenome-TSS1Mb-regulatoryTrackSummary-PCA256-v1.1 + +Embedding ID: `alphagenome_tss1mb_regulatory_tracksummary_pca256` + +Former processing ID: `alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20084 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `embedding_row_index` | 20084 | 20084 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 20084 | 20084 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 20084 | 20057 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 20084 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `chromosome` | 20084 | 24 | chrX
chr20
chr1
chr6 | Chromosome for the gene or modeled interval. | +| `gene_start` | 20084 | 20026 | 100627108
100584936
50934852
169846981 | Gene start coordinate in the stated genome assembly. | +| `gene_end` | 20084 | 19975 | 100639991
100599885
50959140
169894267 | Gene end coordinate in the stated genome assembly. | +| `strand` | 20084 | 2 | -
+ | Genomic strand. | +| `assembly` | 20084 | 1 | GRCh38 | Genome assembly used for coordinates. | +| `tss_1based` | 20084 | 20034 | 100639991
100584936
50959140
169894267 | Transcription start site coordinate in 1-based coordinates. | +| `input_interval` | 20084 | 20050 | chrX:100115702-101164278:-
chrX:100060648-101109224:+
chr20:50434851-51483427:-
chr1:169369978-170418554:- | Genomic interval used as model input. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `embedding_name` | AlphaGenome-TSS1Mb-regulatoryTrackSummary-PCA256-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_embeddings.npz b/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/embeddings.npz similarity index 100% rename from data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_embeddings.npz rename to data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/embeddings.npz diff --git a/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_genes.tsv b/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/genes.tsv similarity index 100% rename from data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_genes.tsv rename to data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/genes.tsv diff --git a/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_metadata.json b/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/metadata.json similarity index 75% rename from data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_metadata.json rename to data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/metadata.json index a32c8c1a03b85e5cefbc2353cc0dda48b2340480..5bf891ca766455bafe4608d26bae28fa8a85000e 100644 --- a/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_metadata.json +++ b/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/metadata.json @@ -30,5 +30,13 @@ "full_1mb_mean", "central_10kb_max" ], - "transform": "signed log1p per summary feature, z-score across genes, PCA to 256 dimensions" -} \ No newline at end of file + "transform": "signed log1p per summary feature, z-score across genes, PCA to 256 dimensions", + "embedding_id": "alphagenome_tss1mb_regulatory_tracksummary_pca256", + "former_processing_embedding_id": "alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/basenji2_tss_131kb_human_central3/README.md b/data/embeddings/basenji2_tss_131kb_human_central3/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e867a1206b872df500dee2be9d0a8e1a16ff158d --- /dev/null +++ b/data/embeddings/basenji2_tss_131kb_human_central3/README.md @@ -0,0 +1,77 @@ +# Basenji2-TSS-131kb-human-central3-log1p-zscore-v1.1 + +Embedding ID: `basenji2_tss_131kb_human_central3` + +Former processing ID: `basenji2_tss_131kb_human_central3_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20097 genes × 5313 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 20097 | 20097 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 20097 | 20070 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `entrez_id` | 19545 | 19511 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 20045 | 19860 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `gene_type` | 20097 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `chromosome` | 20097 | 25 | chrX
chr20
chr1
chr6 | Chromosome for the gene or modeled interval. | +| `gene_start` | 20097 | 20039 | 100627108
100584936
50934852
169846981 | Gene start coordinate in the stated genome assembly. | +| `gene_end` | 20097 | 19988 | 100639991
100599885
50959140
169894267 | Gene end coordinate in the stated genome assembly. | +| `strand` | 20097 | 2 | -
+ | Genomic strand. | +| `tss_1based` | 20097 | 20047 | 100639991
100584936
50959140
169894267 | Transcription start site coordinate in 1-based coordinates. | +| `tss_0based` | 20097 | 20047 | 100639990
100584935
50959139
169894266 | Transcription start site coordinate in 0-based coordinates. | +| `assembly` | 20097 | 1 | GRCh38 | Genome assembly used for coordinates. | +| `gencode_release` | 20097 | 1 | 49 | GENCODE release used for annotation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | regulatory DNA / Basenji2 TSS sequence model | +| `embedding_name` | Basenji2-TSS-131kb-human-central3-log1p-zscore-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_embeddings.npz b/data/embeddings/basenji2_tss_131kb_human_central3/embeddings.npz similarity index 100% rename from data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_embeddings.npz rename to data/embeddings/basenji2_tss_131kb_human_central3/embeddings.npz diff --git a/data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_genes.tsv b/data/embeddings/basenji2_tss_131kb_human_central3/genes.tsv similarity index 100% rename from data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_genes.tsv rename to data/embeddings/basenji2_tss_131kb_human_central3/genes.tsv diff --git a/data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_metadata.json b/data/embeddings/basenji2_tss_131kb_human_central3/metadata.json similarity index 83% rename from data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_metadata.json rename to data/embeddings/basenji2_tss_131kb_human_central3/metadata.json index 891059db2d77c0145bc335dcf008e716d09391c6..8f0ab79331fd7263d037e40f2afb921ccedd759b 100644 --- a/data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_metadata.json +++ b/data/embeddings/basenji2_tss_131kb_human_central3/metadata.json @@ -31,5 +31,13 @@ "n_valid_tss_windows": 20097, "n_skipped_tss_windows": 0, "dimension": 5313, - "created_at": "2026-07-10T06:17:21" -} \ No newline at end of file + "created_at": "2026-07-10T06:17:21", + "embedding_id": "basenji2_tss_131kb_human_central3", + "former_processing_embedding_id": "basenji2_tss_131kb_human_central3_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/biobert_gene_protein_summary/README.md b/data/embeddings/biobert_gene_protein_summary/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8543024bec1966cf21e6b590806d4f383763cc7f --- /dev/null +++ b/data/embeddings/biobert_gene_protein_summary/README.md @@ -0,0 +1,75 @@ +# BioBERT-GeneProteinSummary-v1.1 + +Embedding ID: `biobert_gene_protein_summary` + +Former processing ID: `biobert_gene_protein_summary_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19168 genes × 768 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `row_index` | 19168 | 19168 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 19168 | 19168 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19168 | 19143 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `hgnc_approved_symbol` | 19082 | 19083 | TSPAN6
TNMD
DPM1
SCYL3 | HGNC-approved gene symbol from the master table. | +| `entrez_id` | 19118 | 19092 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19153 | 18980 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `summary_key` | 19168 | 19082 | TSPAN6
TNMD
DPM1
SCYL3 | Identifier used to retrieve the gene/protein text summary. | +| `summary_mapping_source` | 19168 | 3 | gene_symbol
map_symbols_all_fallback
hgnc_approved_symbol | Mapping route used to select the gene/protein text summary. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | literature_text_gene_summary | +| `embedding_name` | BioBERT-GeneProteinSummary-v1.1 | +| `download_source` | GenePT Zenodo v2 summaries | +| `embedding_generated_by` | Daniel locally | +| `model_checkpoint` | dmis-lab/biobert-base-cased-v1.2 | +| `pooling_strategy` | final hidden layer mean pooling over valid non-special, non-padding tokens | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_embeddings.npz b/data/embeddings/biobert_gene_protein_summary/embeddings.npz similarity index 100% rename from data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_embeddings.npz rename to data/embeddings/biobert_gene_protein_summary/embeddings.npz diff --git a/data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_genes.tsv b/data/embeddings/biobert_gene_protein_summary/genes.tsv similarity index 100% rename from data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_genes.tsv rename to data/embeddings/biobert_gene_protein_summary/genes.tsv diff --git a/data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_metadata.json b/data/embeddings/biobert_gene_protein_summary/metadata.json similarity index 75% rename from data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_metadata.json rename to data/embeddings/biobert_gene_protein_summary/metadata.json index 979ae174ac6c375e9de9b6de90bf3352211e7401..5bc5f72b28bdfee43ecaac5a4dd9993c3d983c5c 100644 --- a/data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_metadata.json +++ b/data/embeddings/biobert_gene_protein_summary/metadata.json @@ -23,5 +23,13 @@ "zero_vectors": 0, "nan_count": 0, "inf_count": 0, - "runtime_minutes": 3.477354129155477 -} \ No newline at end of file + "runtime_minutes": 3.477354129155477, + "embedding_id": "biobert_gene_protein_summary", + "former_processing_embedding_id": "biobert_gene_protein_summary_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/bioconceptvec_cbow/README.md b/data/embeddings/bioconceptvec_cbow/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ed282bb8e37441f2758ffd0613814d89117cf82d --- /dev/null +++ b/data/embeddings/bioconceptvec_cbow/README.md @@ -0,0 +1,75 @@ +# BioConceptVec-CBOW-v1.1 + +Embedding ID: `bioconceptvec_cbow` + +Former processing ID: `bioconceptvec_cbow_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `17040 genes × 100 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 17040 | 17040 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 17040 | 17040 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 17040 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 17040 | 17040 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 17039 | 16973 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 17040 | 17040 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 17040 | 17040 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 17040 | 1 | BioConceptVec-CBOW-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 17040 | 1 | entrez_id | Type of source identifier used by the original embedding. | +| `source_key_count` | 17040 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 17040 | 17040 | Gene_7105
Gene_64102
Gene_8813
Gene_57147 | Example source keys mapped to this final gene. | +| `mapping_methods` | 17040 | 1 | entrez_id_from_map_entrez_ids_all | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | biomedical_literature_text | +| `embedding_name` | BioConceptVec-CBOW-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_embeddings.npz b/data/embeddings/bioconceptvec_cbow/embeddings.npz similarity index 100% rename from data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_embeddings.npz rename to data/embeddings/bioconceptvec_cbow/embeddings.npz diff --git a/data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_genes.tsv b/data/embeddings/bioconceptvec_cbow/genes.tsv similarity index 100% rename from data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_genes.tsv rename to data/embeddings/bioconceptvec_cbow/genes.tsv diff --git a/data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_metadata.json b/data/embeddings/bioconceptvec_cbow/metadata.json similarity index 86% rename from data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_metadata.json rename to data/embeddings/bioconceptvec_cbow/metadata.json index 5f6a4418eabfceb0cfb3f8cfe1111bc33e33b1bb..6b6b7ec51140a13414e790cb8719efb198ec30e0 100644 --- a/data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_metadata.json +++ b/data/embeddings/bioconceptvec_cbow/metadata.json @@ -27,5 +27,13 @@ "coverage_of_master_rows": 0.8478877444394686, "n_zero_vectors": 0, "output_npz": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_embeddings.npz", - "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_genes.tsv" -} \ No newline at end of file + "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_genes.tsv", + "embedding_id": "bioconceptvec_cbow", + "former_processing_embedding_id": "bioconceptvec_cbow_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/bioconceptvec_fasttext/README.md b/data/embeddings/bioconceptvec_fasttext/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f85c3d084d4fb75b5c680240df6057eed9ce8f8 --- /dev/null +++ b/data/embeddings/bioconceptvec_fasttext/README.md @@ -0,0 +1,75 @@ +# BioConceptVec-fastText-v1.1 + +Embedding ID: `bioconceptvec_fasttext` + +Former processing ID: `bioconceptvec_fasttext_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `17040 genes × 100 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 17040 | 17040 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 17040 | 17040 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 17040 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 17040 | 17040 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 17039 | 16973 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 17040 | 17040 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 17040 | 17040 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 17040 | 1 | BioConceptVec-fastText-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 17040 | 1 | entrez_id | Type of source identifier used by the original embedding. | +| `source_key_count` | 17040 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 17040 | 17040 | Gene_7105
Gene_64102
Gene_8813
Gene_57147 | Example source keys mapped to this final gene. | +| `mapping_methods` | 17040 | 1 | entrez_id_from_map_entrez_ids_all | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | biomedical_literature_text | +| `embedding_name` | BioConceptVec-fastText-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_embeddings.npz b/data/embeddings/bioconceptvec_fasttext/embeddings.npz similarity index 100% rename from data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_embeddings.npz rename to data/embeddings/bioconceptvec_fasttext/embeddings.npz diff --git a/data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_genes.tsv b/data/embeddings/bioconceptvec_fasttext/genes.tsv similarity index 100% rename from data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_genes.tsv rename to data/embeddings/bioconceptvec_fasttext/genes.tsv diff --git a/data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_metadata.json b/data/embeddings/bioconceptvec_fasttext/metadata.json similarity index 85% rename from data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_metadata.json rename to data/embeddings/bioconceptvec_fasttext/metadata.json index 40d9a82db4714a49b3945029b10114128a3c43db..868f9fa4a8017d27de01db84e880c501725c9d97 100644 --- a/data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_metadata.json +++ b/data/embeddings/bioconceptvec_fasttext/metadata.json @@ -27,5 +27,13 @@ "coverage_of_master_rows": 0.8478877444394686, "n_zero_vectors": 0, "output_npz": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_embeddings.npz", - "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_genes.tsv" -} \ No newline at end of file + "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_genes.tsv", + "embedding_id": "bioconceptvec_fasttext", + "former_processing_embedding_id": "bioconceptvec_fasttext_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/bioconceptvec_glove/README.md b/data/embeddings/bioconceptvec_glove/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9012c025bbf3a87feb4a58f242c8a75c3746bbbe --- /dev/null +++ b/data/embeddings/bioconceptvec_glove/README.md @@ -0,0 +1,75 @@ +# BioConceptVec-GloVe-v1.1 + +Embedding ID: `bioconceptvec_glove` + +Former processing ID: `bioconceptvec_glove_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `17040 genes × 100 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 17040 | 17040 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 17040 | 17040 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 17040 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 17040 | 17040 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 17039 | 16973 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 17040 | 17040 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 17040 | 17040 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 17040 | 1 | BioConceptVec-GloVe-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 17040 | 1 | entrez_id | Type of source identifier used by the original embedding. | +| `source_key_count` | 17040 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 17040 | 17040 | Gene_7105
Gene_64102
Gene_8813
Gene_57147 | Example source keys mapped to this final gene. | +| `mapping_methods` | 17040 | 1 | entrez_id_from_map_entrez_ids_all | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | biomedical_literature_text | +| `embedding_name` | BioConceptVec-GloVe-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_embeddings.npz b/data/embeddings/bioconceptvec_glove/embeddings.npz similarity index 100% rename from data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_embeddings.npz rename to data/embeddings/bioconceptvec_glove/embeddings.npz diff --git a/data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_genes.tsv b/data/embeddings/bioconceptvec_glove/genes.tsv similarity index 100% rename from data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_genes.tsv rename to data/embeddings/bioconceptvec_glove/genes.tsv diff --git a/data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_metadata.json b/data/embeddings/bioconceptvec_glove/metadata.json similarity index 85% rename from data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_metadata.json rename to data/embeddings/bioconceptvec_glove/metadata.json index bd2f2414d74b84f33372dbeddc92c6f709a18e5b..d4c3f54b58a671c4c9c36c9428d325b9788084f2 100644 --- a/data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_metadata.json +++ b/data/embeddings/bioconceptvec_glove/metadata.json @@ -27,5 +27,13 @@ "coverage_of_master_rows": 0.8478877444394686, "n_zero_vectors": 0, "output_npz": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_embeddings.npz", - "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_genes.tsv" -} \ No newline at end of file + "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_genes.tsv", + "embedding_id": "bioconceptvec_glove", + "former_processing_embedding_id": "bioconceptvec_glove_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/bioconceptvec_skipgram/README.md b/data/embeddings/bioconceptvec_skipgram/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c860286fa6fb8a44f15072f7c594a87337f1c899 --- /dev/null +++ b/data/embeddings/bioconceptvec_skipgram/README.md @@ -0,0 +1,75 @@ +# BioConceptVec-SkipGram-v1.1 + +Embedding ID: `bioconceptvec_skipgram` + +Former processing ID: `bioconceptvec_skipgram_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `17040 genes × 100 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 17040 | 17040 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 17040 | 17040 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 17040 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 17040 | 17040 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 17039 | 16973 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 17040 | 17040 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 17040 | 17040 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 17040 | 1 | BioConceptVec-SkipGram-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 17040 | 1 | entrez_id | Type of source identifier used by the original embedding. | +| `source_key_count` | 17040 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 17040 | 17040 | Gene_7105
Gene_64102
Gene_8813
Gene_57147 | Example source keys mapped to this final gene. | +| `mapping_methods` | 17040 | 1 | entrez_id_from_map_entrez_ids_all | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | biomedical_literature_text | +| `embedding_name` | BioConceptVec-SkipGram-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_embeddings.npz b/data/embeddings/bioconceptvec_skipgram/embeddings.npz similarity index 100% rename from data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_embeddings.npz rename to data/embeddings/bioconceptvec_skipgram/embeddings.npz diff --git a/data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_genes.tsv b/data/embeddings/bioconceptvec_skipgram/genes.tsv similarity index 100% rename from data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_genes.tsv rename to data/embeddings/bioconceptvec_skipgram/genes.tsv diff --git a/data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_metadata.json b/data/embeddings/bioconceptvec_skipgram/metadata.json similarity index 85% rename from data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_metadata.json rename to data/embeddings/bioconceptvec_skipgram/metadata.json index 5866860c1786e0c79ef286222c65c2d04ce1813c..95cecbdd878a8643bf47eb5160a3f067dee69551 100644 --- a/data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_metadata.json +++ b/data/embeddings/bioconceptvec_skipgram/metadata.json @@ -27,5 +27,13 @@ "coverage_of_master_rows": 0.8478877444394686, "n_zero_vectors": 0, "output_npz": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_embeddings.npz", - "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_genes.tsv" -} \ No newline at end of file + "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_genes.tsv", + "embedding_id": "bioconceptvec_skipgram", + "former_processing_embedding_id": "bioconceptvec_skipgram_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/README.md b/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/README.md new file mode 100644 index 0000000000000000000000000000000000000000..54d1da4a18b2cafe734b457fe074640756faf89a --- /dev/null +++ b/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/README.md @@ -0,0 +1,73 @@ +# CMap-GSE106127-CGS-mean-landmark978-zscore-v1.1 + +Embedding ID: `cmap_gse106127_cgs_mean_landmark978_zscore` + +Former processing ID: `cmap_gse106127_cgs_mean_landmark978_zscore_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `4070 genes × 978 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `embedding_row` | 4070 | 4070 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 4070 | 4070 | ENSG00000000003
ENSG00000000457
ENSG00000000938
ENSG00000001084 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 4070 | 4063 | TSPAN6
SCYL3
FGR
GCLC | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 4070 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 4070 | 4063 | 7105
57147
2268
2729 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 4070 | 4060 | O43657
Q8IZE3
P09769
P48506 | UniProt accession or identifier, when available. | +| `source_gene_symbol` | 4070 | 4063 | TSPAN6
SCYL3
FGR
GCLC | Original gene symbol from the source before final mapping. | +| `n_cgs_signatures` | 4070 | 6 | 9
1
8
7 | TODO: define this source-specific column during manual curation. | +| `n_cell_lines` | 4070 | 6 | 9
1
8
7 | TODO: define this source-specific column during manual curation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | perturbational transcriptomics / L1000 consensus genetic perturbation signatures | +| `embedding_name` | CMap-GSE106127-CGS-mean-landmark978-zscore-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_embeddings.npz b/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/embeddings.npz similarity index 100% rename from data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_embeddings.npz rename to data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/embeddings.npz diff --git a/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_genes.tsv b/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/genes.tsv similarity index 100% rename from data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_genes.tsv rename to data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/genes.tsv diff --git a/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_metadata.json b/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/metadata.json similarity index 90% rename from data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_metadata.json rename to data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/metadata.json index 84f51a9d2596ec6a8d81852bea03016af321cf97..acd079e26e8212c56eed2a4cbfe20ffcf8f26380 100644 --- a/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_metadata.json +++ b/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/metadata.json @@ -34,5 +34,13 @@ "mapping_report": "/local/home/dtabares/data/processed_embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/mapping_report.txt", "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/unmapped_master_genes.tsv", "unmapped_source_gene_symbols": "/local/home/dtabares/reports/mapping_reports/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/unmapped_source_gene_symbols.tsv" + }, + "embedding_id": "cmap_gse106127_cgs_mean_landmark978_zscore", + "former_processing_embedding_id": "cmap_gse106127_cgs_mean_landmark978_zscore_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/README.md b/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/README.md new file mode 100644 index 0000000000000000000000000000000000000000..351c52f688be486643ed415c403e41b0d573c0dc --- /dev/null +++ b/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/README.md @@ -0,0 +1,79 @@ +# deepNF-STRING-v12-PPMI-SVD-fusion-v1.1 + +Embedding ID: `deepnf_string_v12_ppmi_svd_fusion` + +Former processing ID: `deepnf_string_v12_ppmi_svd_fusion_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19071 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19071 | 19071 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19071 | 19071 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19071 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18908 | 18909 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19060 | 19010 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 19071 | 19071 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 19071 | 19071 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 19071 | 1 | deepNF-STRING-v12-PPMI-SVD-fusion-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 19071 | 1 | STRING protein ID / Ensembl protein ID | Type of source identifier used by the original embedding. | +| `source_key_count` | 19071 | 3 | 1
2
3 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 19071 | 19071 | 9606.ENSP00000362111
9606.ENSP00000362122
9606.ENSP00000360638
9606.ENSP00000356746 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 19071 | 19071 | 10924
10928
10628
9850 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 19071 | 12 | ensembl_protein_id_exact;string_alias_entrez;string_id_exact;string_preferred_symbol
ensembl_protein_id_exact;string_alias_entrez;string_id_exact
ensembl_protein_id_exact;string_alias_entrez;string_id_exact;ensembl_protein_id_exact;string_alias_entrez;string_id_exact;string_preferred_symbol
string_alias_entrez;string_preferred_symbol | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | STRING network multimodal sparse PPMI/SVD fusion | +| `embedding_name` | deepNF-STRING-v12-PPMI-SVD-fusion-v1.1 | +| `download_source` | STRING v12 human protein.links.detailed, protein.info, and protein.aliases files | +| `original_model_or_method` | deepNF: deep network fusion for protein function prediction | +| `embedding_generated_by` | Daniel locally from STRING v12 channel networks | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_embeddings.npz b/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/embeddings.npz similarity index 100% rename from data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_embeddings.npz rename to data/embeddings/deepnf_string_v12_ppmi_svd_fusion/embeddings.npz diff --git a/data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_genes.tsv b/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/genes.tsv similarity index 100% rename from data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_genes.tsv rename to data/embeddings/deepnf_string_v12_ppmi_svd_fusion/genes.tsv diff --git a/data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_metadata.json b/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/metadata.json similarity index 96% rename from data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_metadata.json rename to data/embeddings/deepnf_string_v12_ppmi_svd_fusion/metadata.json index b778d721aff60081d66b834a751cd396814992f4..65c5d7b77e44eb75c39291edadb48e97c374f255 100644 --- a/data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_metadata.json +++ b/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/metadata.json @@ -144,5 +144,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/deepnf_string_v12_ppmi_svd_fusion_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/deepnf_string_v12_ppmi_svd_fusion_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/deepnf_string_v12_ppmi_svd_fusion_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "deepnf_string_v12_ppmi_svd_fusion", + "former_processing_embedding_id": "deepnf_string_v12_ppmi_svd_fusion_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/deepsea_tss_1kb_epigenomic_probs/README.md b/data/embeddings/deepsea_tss_1kb_epigenomic_probs/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4b5be8cf5cab98e9d8d3e3a9865f148f58c2a215 --- /dev/null +++ b/data/embeddings/deepsea_tss_1kb_epigenomic_probs/README.md @@ -0,0 +1,79 @@ +# DeepSEA-TSS-1kb-epigenomic-probabilities-zscore-v1.1 + +Embedding ID: `deepsea_tss_1kb_epigenomic_probs` + +Former processing ID: `deepsea_tss_1kb_epigenomic_probs_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20097 genes × 919 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 20097 | 20097 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 20097 | 20070 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `entrez_id` | 19545 | 19511 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 20045 | 19860 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `gene_type` | 20097 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `chromosome` | 20097 | 25 | chrX
chr20
chr1
chr6 | Chromosome for the gene or modeled interval. | +| `gene_start` | 20097 | 20039 | 100627108
100584936
50934852
169846981 | Gene start coordinate in the stated genome assembly. | +| `gene_end` | 20097 | 19988 | 100639991
100599885
50959140
169894267 | Gene end coordinate in the stated genome assembly. | +| `strand` | 20097 | 2 | -
+ | Genomic strand. | +| `tss_1based` | 20097 | 20047 | 100639991
100584936
50959140
169894267 | Transcription start site coordinate in 1-based coordinates. | +| `bed_start_0based` | 20097 | 20047 | 100639490
100584435
50958639
169893766 | BED-format interval start coordinate, 0-based. | +| `bed_end_0based` | 20097 | 20047 | 100640490
100585435
50959639
169894766 | BED-format interval end coordinate, half-open. | +| `bed_name` | 20097 | 20097 | ENSG00000000003|TSPAN6
ENSG00000000005|TNMD
ENSG00000000419|DPM1
ENSG00000000457|SCYL3 | BED interval name. | +| `assembly` | 20097 | 1 | GRCh38 | Genome assembly used for coordinates. | +| `gencode_release` | 20097 | 1 | 49 | GENCODE release used for annotation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | regulatory DNA / promoter-TSS sequence model | +| `embedding_name` | DeepSEA-TSS-1kb-epigenomic-probabilities-zscore-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_embeddings.npz b/data/embeddings/deepsea_tss_1kb_epigenomic_probs/embeddings.npz similarity index 100% rename from data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_embeddings.npz rename to data/embeddings/deepsea_tss_1kb_epigenomic_probs/embeddings.npz diff --git a/data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_genes.tsv b/data/embeddings/deepsea_tss_1kb_epigenomic_probs/genes.tsv similarity index 100% rename from data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_genes.tsv rename to data/embeddings/deepsea_tss_1kb_epigenomic_probs/genes.tsv diff --git a/data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_metadata.json b/data/embeddings/deepsea_tss_1kb_epigenomic_probs/metadata.json similarity index 83% rename from data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_metadata.json rename to data/embeddings/deepsea_tss_1kb_epigenomic_probs/metadata.json index 690af82b495e297baa697159d468210ffa88bfad..3c7d9a28ce78cbcee2f906d0a66e9d67fb35e9e3 100644 --- a/data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_metadata.json +++ b/data/embeddings/deepsea_tss_1kb_epigenomic_probs/metadata.json @@ -26,5 +26,13 @@ "feature_columns": "deepsea_feature_columns.tsv", "skipped_windows": "/local/home/dtabares/reports/mapping_reports/deepsea_tss_1kb_epigenomic_probs_v1_1/skipped_tss_windows.tsv" }, - "created_at": "2026-07-09T10:41:32" -} \ No newline at end of file + "created_at": "2026-07-09T10:41:32", + "embedding_id": "deepsea_tss_1kb_epigenomic_probs", + "former_processing_embedding_id": "deepsea_tss_1kb_epigenomic_probs_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/depmap_crispr_gene_effect_pca256/README.md b/data/embeddings/depmap_crispr_gene_effect_pca256/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7281d9b382487cfb48acce06836733b7bc4f90d8 --- /dev/null +++ b/data/embeddings/depmap_crispr_gene_effect_pca256/README.md @@ -0,0 +1,73 @@ +# DepMap-CRISPRGeneEffect-PCA256-v1.1 + +Embedding ID: `depmap_crispr_gene_effect_pca256` + +Former processing ID: `depmap_crispr_gene_effect_pca256_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18562 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `embedding_row` | 18562 | 18562 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 18562 | 18562 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18562 | 18537 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18562 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18562 | 18529 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18562 | 18498 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `source_entrez_id` | 18562 | 18529 | 7105
64102
8813
57147 | Entrez Gene identifier from the original source before final mapping. | +| `source_gene_symbol` | 18562 | 18529 | TSPAN6
TNMD
DPM1
SCYL3 | Original gene symbol from the source before final mapping. | +| `source_column` | 18562 | 18529 | TSPAN6 (7105)
TNMD (64102)
DPM1 (8813)
SCYL3 (57147) | TODO: define this source-specific column during manual curation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | functional genomics / CRISPR dependency | +| `embedding_name` | DepMap-CRISPRGeneEffect-PCA256-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_embeddings.npz b/data/embeddings/depmap_crispr_gene_effect_pca256/embeddings.npz similarity index 100% rename from data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_embeddings.npz rename to data/embeddings/depmap_crispr_gene_effect_pca256/embeddings.npz diff --git a/data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_genes.tsv b/data/embeddings/depmap_crispr_gene_effect_pca256/genes.tsv similarity index 100% rename from data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_genes.tsv rename to data/embeddings/depmap_crispr_gene_effect_pca256/genes.tsv diff --git a/data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_metadata.json b/data/embeddings/depmap_crispr_gene_effect_pca256/metadata.json similarity index 88% rename from data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_metadata.json rename to data/embeddings/depmap_crispr_gene_effect_pca256/metadata.json index aecb0a9f59556dc029541da012b0d6239462d009..6d55e3fa885f2d6edcee51c8ad14da0db4cc4692 100644 --- a/data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_metadata.json +++ b/data/embeddings/depmap_crispr_gene_effect_pca256/metadata.json @@ -32,5 +32,13 @@ "metadata": "/local/home/dtabares/data/processed_embeddings/depmap_crispr_gene_effect_pca256_v1_1/metadata.json", "mapping_report": "/local/home/dtabares/data/processed_embeddings/depmap_crispr_gene_effect_pca256_v1_1/mapping_report.txt", "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/depmap_crispr_gene_effect_pca256_v1_1/unmapped_master_genes.tsv" + }, + "embedding_id": "depmap_crispr_gene_effect_pca256", + "former_processing_embedding_id": "depmap_crispr_gene_effect_pca256_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/dnabert2_117m_cds/README.md b/data/embeddings/dnabert2_117m_cds/README.md new file mode 100644 index 0000000000000000000000000000000000000000..466382900a2b81e92758b2eb4978075a0c46e8d2 --- /dev/null +++ b/data/embeddings/dnabert2_117m_cds/README.md @@ -0,0 +1,69 @@ +# dnabert2_117m_cds_v1_1 + +Embedding ID: `dnabert2_117m_cds` + +Former processing ID: `dnabert2_117m_cds_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20085 genes × 768 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 20085 | 20085 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `transcript_id` | 20085 | 20085 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Transcript identifier used for sequence-based processing. | +| `dnabert2_sequence_length` | 20085 | 2255 | 738
954
783
2067 | TODO: define this source-specific column during manual curation. | +| `num_chunks` | 20085 | 19 | 1
2
3
4 | Number of chunks/windows used when the input was too long. | +| `was_chunked` | 20085 | 2 | False
True | Whether the input sequence was split into chunks. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `embedding_name` | dnabert2_117m_cds_v1_1 | +| `pooling_strategy` | Mean pooling over final-layer non-special token embeddings. Long CDS sequences split into nucleotide chunks, then length-weighted mean of chunk-level embeddings. | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_embeddings.npz b/data/embeddings/dnabert2_117m_cds/embeddings.npz similarity index 100% rename from data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_embeddings.npz rename to data/embeddings/dnabert2_117m_cds/embeddings.npz diff --git a/data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_genes.tsv b/data/embeddings/dnabert2_117m_cds/genes.tsv similarity index 100% rename from data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_genes.tsv rename to data/embeddings/dnabert2_117m_cds/genes.tsv diff --git a/data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_metadata.json b/data/embeddings/dnabert2_117m_cds/metadata.json similarity index 85% rename from data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_metadata.json rename to data/embeddings/dnabert2_117m_cds/metadata.json index 1b211cd7ef6e94e34b39e08c23c966206391e2da..7bc80c3c610884e13b498c2a5410ab56789f59e5 100644 --- a/data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_metadata.json +++ b/data/embeddings/dnabert2_117m_cds/metadata.json @@ -31,5 +31,13 @@ "transformers": "4.29.2", "numpy": "1.26.4" }, - "runtime_seconds": 1222.18 -} \ No newline at end of file + "runtime_seconds": 1222.18, + "embedding_id": "dnabert2_117m_cds", + "former_processing_embedding_id": "dnabert2_117m_cds_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/enformer_tss_393kb_human_central3/README.md b/data/embeddings/enformer_tss_393kb_human_central3/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0ae872c44cfab6dae293372264eaa87d3b42599e --- /dev/null +++ b/data/embeddings/enformer_tss_393kb_human_central3/README.md @@ -0,0 +1,77 @@ +# Enformer-TSS-393kb-human-central3-log1p-zscore-v1.1 + +Embedding ID: `enformer_tss_393kb_human_central3` + +Former processing ID: `enformer_tss_393kb_human_central3_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20097 genes × 5313 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 20097 | 20097 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 20097 | 20070 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `entrez_id` | 19545 | 19511 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 20045 | 19860 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `gene_type` | 20097 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `chromosome` | 20097 | 25 | chrX
chr20
chr1
chr6 | Chromosome for the gene or modeled interval. | +| `gene_start` | 20097 | 20039 | 100627108
100584936
50934852
169846981 | Gene start coordinate in the stated genome assembly. | +| `gene_end` | 20097 | 19988 | 100639991
100599885
50959140
169894267 | Gene end coordinate in the stated genome assembly. | +| `strand` | 20097 | 2 | -
+ | Genomic strand. | +| `tss_1based` | 20097 | 20047 | 100639991
100584936
50959140
169894267 | Transcription start site coordinate in 1-based coordinates. | +| `tss_0based` | 20097 | 20047 | 100639990
100584935
50959139
169894266 | Transcription start site coordinate in 0-based coordinates. | +| `assembly` | 20097 | 1 | GRCh38 | Genome assembly used for coordinates. | +| `gencode_release` | 20097 | 1 | 49 | GENCODE release used for annotation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | regulatory DNA / long-range TSS sequence model | +| `embedding_name` | Enformer-TSS-393kb-human-central3-log1p-zscore-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_embeddings.npz b/data/embeddings/enformer_tss_393kb_human_central3/embeddings.npz similarity index 100% rename from data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_embeddings.npz rename to data/embeddings/enformer_tss_393kb_human_central3/embeddings.npz diff --git a/data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_genes.tsv b/data/embeddings/enformer_tss_393kb_human_central3/genes.tsv similarity index 100% rename from data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_genes.tsv rename to data/embeddings/enformer_tss_393kb_human_central3/genes.tsv diff --git a/data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_metadata.json b/data/embeddings/enformer_tss_393kb_human_central3/metadata.json similarity index 82% rename from data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_metadata.json rename to data/embeddings/enformer_tss_393kb_human_central3/metadata.json index 2233292d4bd88a701d7f8cd418fbf7b0c25f1f14..facc8b62a9484969cf1017cd89f09e74c7a7f362 100644 --- a/data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_metadata.json +++ b/data/embeddings/enformer_tss_393kb_human_central3/metadata.json @@ -32,5 +32,13 @@ "n_valid_tss_windows": 20097, "n_skipped_tss_windows": 0, "dimension": 5313, - "created_at": "2026-07-10T06:03:35" -} \ No newline at end of file + "created_at": "2026-07-10T06:03:35", + "embedding_id": "enformer_tss_393kb_human_central3", + "former_processing_embedding_id": "enformer_tss_393kb_human_central3_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/esm2_v1_1/README.md b/data/embeddings/esm2/README.md similarity index 99% rename from data/embeddings/esm2_v1_1/README.md rename to data/embeddings/esm2/README.md index fca0d5d052e20622f8fac5bd372e2405488041cb..dd86585791f37d6c2e814dbeda642ead99e767d4 100644 --- a/data/embeddings/esm2_v1_1/README.md +++ b/data/embeddings/esm2/README.md @@ -46,12 +46,8 @@ This embedding is a gene-level protein-sequence representation derived from ESM2. In this project it was not regenerated locally from raw protein sequences. Instead, the ready gene-level ESM2 table was downloaded from the Zhong gene-embedding benchmark archive and then standardized to the project master gene table. The original representation model is ESM2, a protein language model. The downloaded benchmark representation was reported to use UniProt human protein sequences, the esm2_t33_650M_UR50D checkpoint, and mean pooling over amino-acid token embeddings. The exact UniProt release used by the benchmark source was not recorded locally and should be treated as partially verified provenance. - This embedding should therefore be interpreted as an ESM2 protein-sequence embedding, but with the download source recorded as the Zhong benchmark archive rather than an original local ESM2 inference run. - - ### Provenance fields - download_source: Zhong benchmark archive / Zenodo full_embeddings archive original_model_or_method: ESM2 protein language model embedding_generated_by: Zhong et al. @@ -59,7 +55,6 @@ This embedding should therefore be interpreted as an ESM2 protein-sequence embed input_sequence_source: UniProt human protein sequences, as reported by Zhong model_checkpoint: esm2_t33_650M_UR50D, as reported by Zhong pooling_strategy: mean pooling over amino-acid token embeddings, as reported by Zhong - ### Mapping and filtering The source identifiers were mapped to master_gene_table.csv using exact Entrez ID mapping. Only Entrez IDs that mapped uniquely to one master-table gene were accepted. Source rows that were unmapped or ambiguous across master-table rows were excluded. Duplicate source rows were averaged only when they mapped safely to the same final master gene. diff --git a/data/embeddings/esm2_v1_1/esm2_v1_1_embeddings.npz b/data/embeddings/esm2/embeddings.npz similarity index 100% rename from data/embeddings/esm2_v1_1/esm2_v1_1_embeddings.npz rename to data/embeddings/esm2/embeddings.npz diff --git a/data/embeddings/esm2_v1_1/esm2_v1_1_genes.tsv b/data/embeddings/esm2/genes.tsv similarity index 100% rename from data/embeddings/esm2_v1_1/esm2_v1_1_genes.tsv rename to data/embeddings/esm2/genes.tsv diff --git a/data/embeddings/esm2_v1_1/esm2_v1_1_metadata.json b/data/embeddings/esm2/metadata.json similarity index 89% rename from data/embeddings/esm2_v1_1/esm2_v1_1_metadata.json rename to data/embeddings/esm2/metadata.json index e88948c90aca2de39fd4f49278a11c0a5dc949a3..0e050836e1017406e13f934ed159239e069ce2c2 100644 --- a/data/embeddings/esm2_v1_1/esm2_v1_1_metadata.json +++ b/data/embeddings/esm2/metadata.json @@ -33,5 +33,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/esm2_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/esm2_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/esm2_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "esm2", + "former_processing_embedding_id": "esm2_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/frogs_archs4_256/README.md b/data/embeddings/frogs_archs4_256/README.md new file mode 100644 index 0000000000000000000000000000000000000000..839db8de14a946840cfabe847437a19db4ee2cba --- /dev/null +++ b/data/embeddings/frogs_archs4_256/README.md @@ -0,0 +1,68 @@ +# frogs_archs4_256_v1_1 + +Embedding ID: `frogs_archs4_256` + +Former processing ID: `frogs_archs4_256_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `16265 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 16265 | 16265 | ENSG00000000005
ENSG00000000457
ENSG00000000938
ENSG00000000971 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 16265 | 16265 | TNMD
SCYL3
FGR
CFH | Gene symbol associated with the Ensembl gene ID. | +| `source_entrez_id` | 16265 | 16265 | 64102
57147
2268
3075 | Entrez Gene identifier from the original source before final mapping. | +| `master_entrez_id` | 16264 | 16265 | 64102
57147
2268
3075 | Entrez Gene identifier from the master table. | +| `uniprot_id` | 16260 | 16152 | Q9H2S6
Q8IZE3
P09769
P08603 | UniProt accession or identifier, when available. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `embedding_name` | frogs_archs4_256_v1_1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_embeddings.npz b/data/embeddings/frogs_archs4_256/embeddings.npz similarity index 100% rename from data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_embeddings.npz rename to data/embeddings/frogs_archs4_256/embeddings.npz diff --git a/data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_genes.tsv b/data/embeddings/frogs_archs4_256/genes.tsv similarity index 100% rename from data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_genes.tsv rename to data/embeddings/frogs_archs4_256/genes.tsv diff --git a/data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_metadata.json b/data/embeddings/frogs_archs4_256/metadata.json similarity index 83% rename from data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_metadata.json rename to data/embeddings/frogs_archs4_256/metadata.json index dacaae3d8006f376f96b8746f9446e90a0fa307b..6490078ec48d00bdef36a7eb9363a6e49450ec5a 100644 --- a/data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_metadata.json +++ b/data/embeddings/frogs_archs4_256/metadata.json @@ -30,5 +30,13 @@ "duplicate_entrez_rows_before_averaging": 0, "duplicated_ensembl_rows_before_final_collapse": 6, "ambiguous_entrez_ids_excluded_from_master_mapping": 272, - "runtime_seconds": 3.64 -} \ No newline at end of file + "runtime_seconds": 3.64, + "embedding_id": "frogs_archs4_256", + "former_processing_embedding_id": "frogs_archs4_256_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/gene2vec/README.md b/data/embeddings/gene2vec/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c59b2cb81e533cb1e5696348c0e81d6faae161d3 --- /dev/null +++ b/data/embeddings/gene2vec/README.md @@ -0,0 +1,75 @@ +# Gene2Vec-v1.1 + +Embedding ID: `gene2vec` + +Former processing ID: `gene2vec_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `17916 genes × 200 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 17916 | 17916 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 17916 | 17916 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 17916 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 17915 | 17915 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 17912 | 17780 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 17916 | 17916 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 17916 | 17916 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 17916 | 1 | Gene2Vec-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 17916 | 1 | gene_symbol | Type of source identifier used by the original embedding. | +| `source_key_count` | 17916 | 3 | 1
2
3 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 17916 | 17916 | TSPAN6
TNMD
DPM1
SCYL3 | Example source keys mapped to this final gene. | +| `mapping_methods` | 17916 | 5 | direct_current_gene_symbol
hgnc_previous_symbol
hgnc_alias_symbol
direct_current_gene_symbol;hgnc_previous_symbol | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | coexpression_expression | +| `embedding_name` | Gene2Vec-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/gene2vec_v1_1/gene2vec_v1_1_embeddings.npz b/data/embeddings/gene2vec/embeddings.npz similarity index 100% rename from data/embeddings/gene2vec_v1_1/gene2vec_v1_1_embeddings.npz rename to data/embeddings/gene2vec/embeddings.npz diff --git a/data/embeddings/gene2vec_v1_1/gene2vec_v1_1_genes.tsv b/data/embeddings/gene2vec/genes.tsv similarity index 100% rename from data/embeddings/gene2vec_v1_1/gene2vec_v1_1_genes.tsv rename to data/embeddings/gene2vec/genes.tsv diff --git a/data/embeddings/gene2vec_v1_1/gene2vec_v1_1_metadata.json b/data/embeddings/gene2vec/metadata.json similarity index 82% rename from data/embeddings/gene2vec_v1_1/gene2vec_v1_1_metadata.json rename to data/embeddings/gene2vec/metadata.json index 7092aacc4589e232db9d5f4d5f996b00881fa118..697e5502f9c97df93b91120318da0041a1373cb1 100644 --- a/data/embeddings/gene2vec_v1_1/gene2vec_v1_1_metadata.json +++ b/data/embeddings/gene2vec/metadata.json @@ -26,5 +26,13 @@ "coverage_of_master_rows": 0.8914763397522019, "n_zero_vectors": 0, "output_npz": "/local/home/dtabares/data/processed_embeddings/gene2vec_v1_1/gene2vec_v1_1_embeddings.npz", - "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/gene2vec_v1_1/gene2vec_v1_1_genes.tsv" -} \ No newline at end of file + "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/gene2vec_v1_1/gene2vec_v1_1_genes.tsv", + "embedding_id": "gene2vec", + "former_processing_embedding_id": "gene2vec_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/genecompass_base_gene_token/README.md b/data/embeddings/genecompass_base_gene_token/README.md new file mode 100644 index 0000000000000000000000000000000000000000..563d390baf86d1dde0bf3d90276fc0078117d19b --- /dev/null +++ b/data/embeddings/genecompass_base_gene_token/README.md @@ -0,0 +1,75 @@ +# genecompass_base_gene_token_v1_1 + +Embedding ID: `genecompass_base_gene_token` + +Former processing ID: `genecompass_base_gene_token_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `16578 genes × 768 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `master_row_index` | 16578 | 16578 | 0
1
2
3 | Row index of the corresponding gene in the master gene table. | +| `ensembl_gene_id` | 16578 | 16578 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 16578 | 16578 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `entrez_id` | 16556 | 16557 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 16565 | 16474 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `genecompass_token_id` | 16578 | 16578 | 16789
13066
11603
22048 | TODO: define this source-specific column during manual curation. | +| `status` | 16578 | 1 | mapped_exact_ensembl_gene_id | Processing or QC status for this row. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | gene expression single-cell foundation model / knowledge-informed gene-token embedding | +| `embedding_name` | genecompass_base_gene_token_v1_1 | +| `download_source` | official GeneCompass ScienceDB checkpoint linked from official GitHub README | +| `original_model_or_method` | GeneCompass Base, 12-layer knowledge-informed cross-species foundation model | +| `embedding_generated_by` | Daniel locally by extracting checkpoint embedding weights; no training or inference | +| `pooling_strategy` | none; static learned token embedding weight extracted directly from checkpoint | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_embeddings.npz b/data/embeddings/genecompass_base_gene_token/embeddings.npz similarity index 100% rename from data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_embeddings.npz rename to data/embeddings/genecompass_base_gene_token/embeddings.npz diff --git a/data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_genes.tsv b/data/embeddings/genecompass_base_gene_token/genes.tsv similarity index 100% rename from data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_genes.tsv rename to data/embeddings/genecompass_base_gene_token/genes.tsv diff --git a/data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_metadata.json b/data/embeddings/genecompass_base_gene_token/metadata.json similarity index 92% rename from data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_metadata.json rename to data/embeddings/genecompass_base_gene_token/metadata.json index 2aa3cd61c19ee545b986c3c910c90c8efd1e423a..8a7420c7940f9eee8b4b1c17947a575eac4b1b65 100644 --- a/data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_metadata.json +++ b/data/embeddings/genecompass_base_gene_token/metadata.json @@ -65,5 +65,13 @@ "status_counts": { "mapped_exact_ensembl_gene_id": 16578, "missing_genecompass_token": 3519 + }, + "embedding_id": "genecompass_base_gene_token", + "former_processing_embedding_id": "genecompass_base_gene_token_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/geneformer_v2_316m/README.md b/data/embeddings/geneformer_v2_316m/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e285d2d0ccaa1a68f4c5c70c52eebc24bb5ee9e6 --- /dev/null +++ b/data/embeddings/geneformer_v2_316m/README.md @@ -0,0 +1,78 @@ +# Geneformer-V2-316M-v1.1 + +Embedding ID: `geneformer_v2_316m` + +Former processing ID: `geneformer_v2_316m_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20007 genes × 1152 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 20007 | 20007 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 20007 | 20007 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 20007 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19488 | 19481 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19964 | 19804 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 20007 | 20007 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 20007 | 20007 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 20007 | 1 | ctheodoris_Geneformer_default_model | Name of the original source embedding. | +| `source_identifier_type` | 20007 | 1 | Geneformer token dictionary Ensembl gene ID | Type of source identifier used by the original embedding. | +| `source_key_count` | 20007 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 20007 | 20007 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Example source keys mapped to this final gene. | +| `source_token_ids_examples` | 20007 | 20007 | 4
5
6
7 | Example source token IDs mapped to this final gene. | +| `mapping_methods` | 20007 | 1 | ensembl_gene_id_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | single_cell_expression_foundation_model | +| `embedding_name` | Geneformer-V2-316M-v1.1 | +| `download_source` | Official ctheodoris/Geneformer Hugging Face repository | +| `embedding_generated_by` | Extracted locally from official Geneformer model input gene-token embedding layer | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_embeddings.npz b/data/embeddings/geneformer_v2_316m/embeddings.npz similarity index 100% rename from data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_embeddings.npz rename to data/embeddings/geneformer_v2_316m/embeddings.npz diff --git a/data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_genes.tsv b/data/embeddings/geneformer_v2_316m/genes.tsv similarity index 100% rename from data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_genes.tsv rename to data/embeddings/geneformer_v2_316m/genes.tsv diff --git a/data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_metadata.json b/data/embeddings/geneformer_v2_316m/metadata.json similarity index 91% rename from data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_metadata.json rename to data/embeddings/geneformer_v2_316m/metadata.json index 4bc3dbdc2665f1d384cf445455d1b24333a4f53e..e5bdf222b2a9a402c2eb27f218d0840428569cbe 100644 --- a/data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_metadata.json +++ b/data/embeddings/geneformer_v2_316m/metadata.json @@ -39,5 +39,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/geneformer_v2_316m_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/geneformer_v2_316m_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/geneformer_v2_316m_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "geneformer_v2_316m", + "former_processing_embedding_id": "geneformer_v2_316m_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/genept_ada_conservative/README.md b/data/embeddings/genept_ada_conservative/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d45cf16f59ad90f793530ae166c545a8d66b7c73 --- /dev/null +++ b/data/embeddings/genept_ada_conservative/README.md @@ -0,0 +1,75 @@ +# GenePT-ADA-v1.1-conservative + +Embedding ID: `genept_ada_conservative` + +Former processing ID: `genept_ada_v1_1_conservative` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19099 genes × 1536 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19099 | 19099 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19099 | 19099 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19099 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19098 | 19099 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19095 | 18945 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 19099 | 19099 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 19099 | 19099 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 19099 | 1 | GenePT-ADA-v1.1-conservative | Name of the original source embedding. | +| `source_identifier_type` | 19099 | 1 | gene_symbol | Type of source identifier used by the original embedding. | +| `source_key_count` | 19099 | 6 | 1
3
2
5 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 19099 | 19099 | TSPAN6
TNMD
DPM1
SCYL3 | Example source keys mapped to this final gene. | +| `mapping_methods` | 19099 | 3 | direct_current_gene_symbol
hgnc_previous_symbol
hgnc_alias_symbol | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | biomedical_literature_text | +| `embedding_name` | GenePT-ADA-v1.1-conservative | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_embeddings.npz b/data/embeddings/genept_ada_conservative/embeddings.npz similarity index 100% rename from data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_embeddings.npz rename to data/embeddings/genept_ada_conservative/embeddings.npz diff --git a/data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_genes.tsv b/data/embeddings/genept_ada_conservative/genes.tsv similarity index 100% rename from data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_genes.tsv rename to data/embeddings/genept_ada_conservative/genes.tsv diff --git a/data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_metadata.json b/data/embeddings/genept_ada_conservative/metadata.json similarity index 86% rename from data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_metadata.json rename to data/embeddings/genept_ada_conservative/metadata.json index 7043ccb737e7cf722ab699dc4159a9a1f1e78cda..57a125a3f88aa0d68c72ac8d29c1c1f14cd057e6 100644 --- a/data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_metadata.json +++ b/data/embeddings/genept_ada_conservative/metadata.json @@ -30,5 +30,13 @@ "coverage_of_master_rows": 0.950340846892571, "n_zero_vectors": 0, "output_npz": "/local/home/dtabares/data/processed_embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_embeddings.npz", - "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_genes.tsv" -} \ No newline at end of file + "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_genes.tsv", + "embedding_id": "genept_ada_conservative", + "former_processing_embedding_id": "genept_ada_v1_1_conservative", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/genept_model3_gene_protein_conservative/README.md b/data/embeddings/genept_model3_gene_protein_conservative/README.md new file mode 100644 index 0000000000000000000000000000000000000000..323dcf4bc8e431391e2f29ca61cb00328f6a425b --- /dev/null +++ b/data/embeddings/genept_model3_gene_protein_conservative/README.md @@ -0,0 +1,75 @@ +# GenePT-Model3-GeneProtein-v1.1-conservative + +Embedding ID: `genept_model3_gene_protein_conservative` + +Former processing ID: `genept_model3_gene_protein_v1_1_conservative` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19103 genes × 3072 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19103 | 19103 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19103 | 19103 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19103 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19102 | 19103 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19099 | 18949 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 19103 | 19103 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 19103 | 19103 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 19103 | 1 | GenePT-Model3-GeneProtein-v1.1-conservative | Name of the original source embedding. | +| `source_identifier_type` | 19103 | 1 | gene_symbol | Type of source identifier used by the original embedding. | +| `source_key_count` | 19103 | 8 | 1
2
5
3 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 19103 | 19103 | TSPAN6
TNMD
DPM1
SCYL3; SCYL3 | Example source keys mapped to this final gene. | +| `mapping_methods` | 19103 | 3 | direct_current_gene_symbol
hgnc_previous_symbol
hgnc_alias_symbol | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | biomedical_literature_text_plus_protein_sequence_text | +| `embedding_name` | GenePT-Model3-GeneProtein-v1.1-conservative | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_embeddings.npz b/data/embeddings/genept_model3_gene_protein_conservative/embeddings.npz similarity index 100% rename from data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_embeddings.npz rename to data/embeddings/genept_model3_gene_protein_conservative/embeddings.npz diff --git a/data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_genes.tsv b/data/embeddings/genept_model3_gene_protein_conservative/genes.tsv similarity index 100% rename from data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_genes.tsv rename to data/embeddings/genept_model3_gene_protein_conservative/genes.tsv diff --git a/data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_metadata.json b/data/embeddings/genept_model3_gene_protein_conservative/metadata.json similarity index 84% rename from data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_metadata.json rename to data/embeddings/genept_model3_gene_protein_conservative/metadata.json index d7122e152ad9b1c5fe6f82b635c9f5bc7646ef3c..25dbb11128299cae6c3bf15ddfc9664f7e52259f 100644 --- a/data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_metadata.json +++ b/data/embeddings/genept_model3_gene_protein_conservative/metadata.json @@ -30,5 +30,13 @@ "coverage_of_master_rows": 0.9505398815743643, "n_zero_vectors": 0, "output_npz": "/local/home/dtabares/data/processed_embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_embeddings.npz", - "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_genes.tsv" -} \ No newline at end of file + "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_genes.tsv", + "embedding_id": "genept_model3_gene_protein_conservative", + "former_processing_embedding_id": "genept_model3_gene_protein_v1_1_conservative", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/gtex_tissue_median_log1p_zscore/README.md b/data/embeddings/gtex_tissue_median_log1p_zscore/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ce41273c1dfd01bc83af134fadb2dd7b6c1a540a --- /dev/null +++ b/data/embeddings/gtex_tissue_median_log1p_zscore/README.md @@ -0,0 +1,73 @@ +# GTEx-tissue-median-log1p-zscore-v1.1 + +Embedding ID: `gtex_tissue_median_log1p_zscore` + +Former processing ID: `gtex_tissue_median_log1p_zscore_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19353 genes × 68 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `embedding_row` | 19353 | 19353 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 19353 | 19353 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19353 | 19335 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19353 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19320 | 19302 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19349 | 19201 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `source_ensembl_gene_id` | 19353 | 19353 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Original Ensembl gene identifier from the source. | +| `source_ensembl_gene_id_versioned` | 19353 | 19353 | ENSG00000000003.16
ENSG00000000005.6
ENSG00000000419.14
ENSG00000000457.14 | Original versioned Ensembl gene identifier from the source. | +| `source_gene_symbol` | 19353 | 19335 | TSPAN6
TNMD
DPM1
SCYL3 | Original gene symbol from the source before final mapping. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | bulk RNA expression / normal tissue expression | +| `embedding_name` | GTEx-tissue-median-log1p-zscore-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_embeddings.npz b/data/embeddings/gtex_tissue_median_log1p_zscore/embeddings.npz similarity index 100% rename from data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_embeddings.npz rename to data/embeddings/gtex_tissue_median_log1p_zscore/embeddings.npz diff --git a/data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_genes.tsv b/data/embeddings/gtex_tissue_median_log1p_zscore/genes.tsv similarity index 100% rename from data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_genes.tsv rename to data/embeddings/gtex_tissue_median_log1p_zscore/genes.tsv diff --git a/data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_metadata.json b/data/embeddings/gtex_tissue_median_log1p_zscore/metadata.json similarity index 86% rename from data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_metadata.json rename to data/embeddings/gtex_tissue_median_log1p_zscore/metadata.json index 88742c7eb0e04e480ffca4e7cd19a4a6f115c274..e6417709540d4c995408fc932563f533ab2d080a 100644 --- a/data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_metadata.json +++ b/data/embeddings/gtex_tissue_median_log1p_zscore/metadata.json @@ -26,5 +26,13 @@ "metadata": "/local/home/dtabares/data/processed_embeddings/gtex_tissue_median_log1p_zscore_v1_1/metadata.json", "mapping_report": "/local/home/dtabares/data/processed_embeddings/gtex_tissue_median_log1p_zscore_v1_1/mapping_report.txt", "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/gtex_tissue_median_log1p_zscore_v1_1/unmapped_master_genes.tsv" + }, + "embedding_id": "gtex_tissue_median_log1p_zscore", + "former_processing_embedding_id": "gtex_tissue_median_log1p_zscore_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/helix_mrna_full_transcript/README.md b/data/embeddings/helix_mrna_full_transcript/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8e3659d70685baf0ea5e3da2794090a40247bd63 --- /dev/null +++ b/data/embeddings/helix_mrna_full_transcript/README.md @@ -0,0 +1,85 @@ +# helix_mrna_full_transcript_v1_1 + +Embedding ID: `helix_mrna_full_transcript` + +Former processing ID: `helix_mrna_full_transcript_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19976 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `master_row_index` | 19976 | 19976 | 0
1
2
3 | Row index of the corresponding gene in the master gene table. | +| `ensembl_gene_id` | 19976 | 19976 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19976 | 19959 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `entrez_id` | 19526 | 19503 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19928 | 19752 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 19976 | 19976 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_transcript_id_versioned` | 19976 | 19976 | ENST00000373020.9
ENST00000373031.5
ENST00000371588.10
ENST00000367771.11 | Versioned canonical transcript identifier. | +| `source_transcript_id_used` | 19976 | 19976 | ENST00000373020.9
ENST00000373031.5
ENST00000371588.10
ENST00000367771.11 | Original transcript identifier selected as model input. | +| `mrna_length_nt` | 19976 | 7597 | 3768
1205
1054
6308 | mRNA/transcript sequence length in nucleotides. | +| `cds_length_nt` | 19976 | 2238 | 738
954
783
2067 | Coding sequence length in nucleotides. | +| `cds_start_0based` | 19976 | 1107 | 112
83
9
159 | TODO: define this source-specific column during manual curation. | +| `cds_end_0based_exclusive` | 19976 | 4794 | 850
1037
792
2226 | TODO: define this source-specific column during manual curation. | +| `helix_input_length` | 19976 | 8206 | 4014
1523
1315
6997 | TODO: define this source-specific column during manual curation. | +| `mapping_method` | 19976 | 2 | canonical_transcript_id_versioned_exact__cds_substring_match
canonical_transcript_id_unversioned_exact__cds_substring_match | Mapping method used to connect the source record to the master gene table. | +| `helix_token_count_used_for_mean_pooling` | 19976 | 8206 | 4014
1523
1315
6997 | TODO: define this source-specific column during manual curation. | +| `helix_chunk_count` | 19976 | 65 | 8
3
14
10 | TODO: define this source-specific column during manual curation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | mRNA language model / mature transcript sequence | +| `embedding_name` | helix_mrna_full_transcript_v1_1 | +| `download_source` | official HuggingFace model repository | +| `original_model_or_method` | Helix-mRNA | +| `embedding_generated_by` | Daniel locally by model inference | +| `model_checkpoint` | helical-ai/helix-mRNA | +| `pooling_strategy` | final token embeddings from official Helix wrapper, mean pooled over non-special sequence tokens; long transcripts chunked and token-count weighted | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_embeddings.npz b/data/embeddings/helix_mrna_full_transcript/embeddings.npz similarity index 100% rename from data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_embeddings.npz rename to data/embeddings/helix_mrna_full_transcript/embeddings.npz diff --git a/data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_genes.tsv b/data/embeddings/helix_mrna_full_transcript/genes.tsv similarity index 100% rename from data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_genes.tsv rename to data/embeddings/helix_mrna_full_transcript/genes.tsv diff --git a/data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_metadata.json b/data/embeddings/helix_mrna_full_transcript/metadata.json similarity index 90% rename from data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_metadata.json rename to data/embeddings/helix_mrna_full_transcript/metadata.json index 350c2f398403d4cd1f27b0be50fa9edf439e6edb..df0ade2efc9e02ce2d3c96cc57d12597ece78018 100644 --- a/data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_metadata.json +++ b/data/embeddings/helix_mrna_full_transcript/metadata.json @@ -47,5 +47,13 @@ "The official helical HelixmRNA wrapper was used instead.", "121 master-table genes did not have a clean canonical cDNA/CDS match and were excluded from the processed matrix.", "Fast Mamba kernels were unavailable; the official wrapper used the naive implementation." - ] -} \ No newline at end of file + ], + "embedding_id": "helix_mrna_full_transcript", + "former_processing_embedding_id": "helix_mrna_full_transcript_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/README.md b/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4a66e33a5cdcadd5e654cf77f5744678c3e2d760 --- /dev/null +++ b/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/README.md @@ -0,0 +1,72 @@ +# Hetionet-nonPPI-biomedical-context-TFIDF-SVD256-v1.1 + +Embedding ID: `hetionet_nonppi_biomedical_context_tfidf_svd256` + +Former processing ID: `hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18787 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `embedding_row` | 18787 | 18787 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 18787 | 18787 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18787 | 18763 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18787 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18787 | 18755 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18786 | 18634 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `source_entrez_id` | 18787 | 18755 | 7105
64102
8813
57147 | Entrez Gene identifier from the original source before final mapping. | +| `source_gene_symbol` | 18787 | 18755 | TSPAN6
TNMD
DPM1
SCYL3 | Original gene symbol from the source before final mapping. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | heterogeneous biomedical knowledge graph / non-PPI gene context | +| `embedding_name` | Hetionet-nonPPI-biomedical-context-TFIDF-SVD256-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_embeddings.npz b/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/embeddings.npz similarity index 100% rename from data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_embeddings.npz rename to data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/embeddings.npz diff --git a/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_genes.tsv b/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/genes.tsv similarity index 100% rename from data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_genes.tsv rename to data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/genes.tsv diff --git a/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_metadata.json b/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/metadata.json similarity index 91% rename from data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_metadata.json rename to data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/metadata.json index ed8b831e71825b51cf207f5ce847ccccfcb39b5d..de8b459a94cae77122b8e04a79acdc689aaa2771 100644 --- a/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_metadata.json +++ b/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/metadata.json @@ -45,5 +45,13 @@ "mapping_report": "/local/home/dtabares/data/processed_embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/mapping_report.txt", "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/unmapped_master_genes.tsv", "unmapped_source_genes": "/local/home/dtabares/reports/mapping_reports/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/unmapped_source_genes.tsv" + }, + "embedding_id": "hetionet_nonppi_biomedical_context_tfidf_svd256", + "former_processing_embedding_id": "hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/hpa_normal_ihc_tissue_level_zscore/README.md b/data/embeddings/hpa_normal_ihc_tissue_level_zscore/README.md new file mode 100644 index 0000000000000000000000000000000000000000..593fca5fa663996db7d6bcff05218a8852b93f79 --- /dev/null +++ b/data/embeddings/hpa_normal_ihc_tissue_level_zscore/README.md @@ -0,0 +1,72 @@ +# HPA-normal-IHC-tissue-level-zscore-v1.1 + +Embedding ID: `hpa_normal_ihc_tissue_level_zscore` + +Former processing ID: `hpa_normal_ihc_tissue_level_zscore_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `13420 genes × 59 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `embedding_row` | 13420 | 13420 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 13420 | 13420 | ENSG00000000003
ENSG00000000419
ENSG00000000457
ENSG00000000460 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 13420 | 13417 | TSPAN6
DPM1
SCYL3
FIRRM | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 13420 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 13372 | 13370 | 7105
8813
57147
55732 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 13418 | 13321 | O43657
O60762
Q8IZE3
Q9NSG2 | UniProt accession or identifier, when available. | +| `hpa_gene` | 13420 | 13420 | ENSG00000000003
ENSG00000000419
ENSG00000000457
ENSG00000000460 | TODO: define this source-specific column during manual curation. | +| `hpa_gene_name` | 13420 | 13416 | TSPAN6
DPM1
SCYL3
C1orf112 | TODO: define this source-specific column during manual curation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | protein expression / normal tissue immunohistochemistry | +| `embedding_name` | HPA-normal-IHC-tissue-level-zscore-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_embeddings.npz b/data/embeddings/hpa_normal_ihc_tissue_level_zscore/embeddings.npz similarity index 100% rename from data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_embeddings.npz rename to data/embeddings/hpa_normal_ihc_tissue_level_zscore/embeddings.npz diff --git a/data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_genes.tsv b/data/embeddings/hpa_normal_ihc_tissue_level_zscore/genes.tsv similarity index 100% rename from data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_genes.tsv rename to data/embeddings/hpa_normal_ihc_tissue_level_zscore/genes.tsv diff --git a/data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_metadata.json b/data/embeddings/hpa_normal_ihc_tissue_level_zscore/metadata.json similarity index 88% rename from data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_metadata.json rename to data/embeddings/hpa_normal_ihc_tissue_level_zscore/metadata.json index 25ccf8097de1fd596a232ea65cb6b0a703f579bb..f3ffb2d9190e187716eda8c64add752de99c0082 100644 --- a/data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_metadata.json +++ b/data/embeddings/hpa_normal_ihc_tissue_level_zscore/metadata.json @@ -51,5 +51,13 @@ "metadata": "/local/home/dtabares/data/processed_embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/metadata.json", "mapping_report": "/local/home/dtabares/data/processed_embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/mapping_report.txt", "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/hpa_normal_ihc_tissue_level_zscore_v1_1/unmapped_master_genes.tsv" + }, + "embedding_id": "hpa_normal_ihc_tissue_level_zscore", + "former_processing_embedding_id": "hpa_normal_ihc_tissue_level_zscore_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/hpa_subcellular_location_multihot_zscore/README.md b/data/embeddings/hpa_subcellular_location_multihot_zscore/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c3b9fec1b280873c90a080fe99072eef5618bb2a --- /dev/null +++ b/data/embeddings/hpa_subcellular_location_multihot_zscore/README.md @@ -0,0 +1,73 @@ +# HPA-subcellular-location-multihot-zscore-v1.1 + +Embedding ID: `hpa_subcellular_location_multihot_zscore` + +Former processing ID: `hpa_subcellular_location_multihot_zscore_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `13581 genes × 50 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `embedding_row` | 13581 | 13581 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 13581 | 13581 | ENSG00000000003
ENSG00000000457
ENSG00000000460
ENSG00000000938 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 13581 | 13575 | TSPAN6
SCYL3
FIRRM
FGR | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 13581 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 13487 | 13482 | 7105
57147
55732
2268 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 13571 | 13485 | O43657
Q8IZE3
Q9NSG2
P09769 | UniProt accession or identifier, when available. | +| `hpa_gene` | 13581 | 13581 | ENSG00000000003
ENSG00000000457
ENSG00000000460
ENSG00000000938 | TODO: define this source-specific column during manual curation. | +| `hpa_gene_name` | 13581 | 13577 | TSPAN6
SCYL3
C1orf112
FGR | TODO: define this source-specific column during manual curation. | +| `hpa_reliability` | 13581 | 4 | Approved
Supported
Uncertain
Enhanced | TODO: define this source-specific column during manual curation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | protein subcellular localization | +| `embedding_name` | HPA-subcellular-location-multihot-zscore-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_embeddings.npz b/data/embeddings/hpa_subcellular_location_multihot_zscore/embeddings.npz similarity index 100% rename from data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_embeddings.npz rename to data/embeddings/hpa_subcellular_location_multihot_zscore/embeddings.npz diff --git a/data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_genes.tsv b/data/embeddings/hpa_subcellular_location_multihot_zscore/genes.tsv similarity index 100% rename from data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_genes.tsv rename to data/embeddings/hpa_subcellular_location_multihot_zscore/genes.tsv diff --git a/data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_metadata.json b/data/embeddings/hpa_subcellular_location_multihot_zscore/metadata.json similarity index 86% rename from data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_metadata.json rename to data/embeddings/hpa_subcellular_location_multihot_zscore/metadata.json index 0e06c5b209850ab2d8f1941427d55725332a7ee3..404f35d68ecc9d3d00b758dfe681cc90fd6ec09d 100644 --- a/data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_metadata.json +++ b/data/embeddings/hpa_subcellular_location_multihot_zscore/metadata.json @@ -32,5 +32,13 @@ "metadata": "/local/home/dtabares/data/processed_embeddings/hpa_subcellular_location_multihot_zscore_v1_1/metadata.json", "mapping_report": "/local/home/dtabares/data/processed_embeddings/hpa_subcellular_location_multihot_zscore_v1_1/mapping_report.txt", "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/hpa_subcellular_location_multihot_zscore_v1_1/unmapped_master_genes.tsv" + }, + "embedding_id": "hpa_subcellular_location_multihot_zscore", + "former_processing_embedding_id": "hpa_subcellular_location_multihot_zscore_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/hyenadna_medium_160k_cds/README.md b/data/embeddings/hyenadna_medium_160k_cds/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9992a459eb801911881af2b78fdf553243b9dd0e --- /dev/null +++ b/data/embeddings/hyenadna_medium_160k_cds/README.md @@ -0,0 +1,81 @@ +# hyenadna_medium_160k_cds_v1_1 + +Embedding ID: `hyenadna_medium_160k_cds` + +Former processing ID: `hyenadna_medium_160k_cds_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20085 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `master_row_index` | 20085 | 20085 | 0
1
2
3 | Row index of the corresponding gene in the master gene table. | +| `ensembl_gene_id` | 20085 | 20085 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 20085 | 20068 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `entrez_id` | 19535 | 19511 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 20036 | 19860 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 20085 | 20085 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_transcript_id_versioned` | 20085 | 20085 | ENST00000373020.9
ENST00000373031.5
ENST00000371588.10
ENST00000367771.11 | Versioned canonical transcript identifier. | +| `source_transcript_id_used` | 20085 | 20085 | ENST00000373020.9
ENST00000373031.5
ENST00000371588.10
ENST00000367771.11 | Original transcript identifier selected as model input. | +| `cds_length_nt` | 20085 | 2255 | 738
954
783
2067 | Coding sequence length in nucleotides. | +| `mapping_method` | 20085 | 2 | canonical_transcript_id_versioned_exact
canonical_transcript_id_unversioned_exact | Mapping method used to connect the source record to the master gene table. | +| `status` | 20085 | 1 | mapped | Processing or QC status for this row. | +| `hyenadna_token_count_used_for_mean_pooling` | 20085 | 2255 | 738
954
783
2067 | TODO: define this source-specific column during manual curation. | +| `hyenadna_chunk_count` | 20085 | 5 | 1
2
3
9 | TODO: define this source-specific column during manual curation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | DNA language model / canonical CDS sequence | +| `embedding_name` | hyenadna_medium_160k_cds_v1_1 | +| `download_source` | official HuggingFace repository LongSafari/hyenadna-medium-160k-seqlen-hf | +| `original_model_or_method` | HyenaDNA medium 160k sequence-length checkpoint | +| `embedding_generated_by` | Daniel locally by model inference | +| `pooling_strategy` | final hidden-state mean pooling over non-special nucleotide tokens; long CDS split into chunks and token-count weighted | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_embeddings.npz b/data/embeddings/hyenadna_medium_160k_cds/embeddings.npz similarity index 100% rename from data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_embeddings.npz rename to data/embeddings/hyenadna_medium_160k_cds/embeddings.npz diff --git a/data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_genes.tsv b/data/embeddings/hyenadna_medium_160k_cds/genes.tsv similarity index 100% rename from data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_genes.tsv rename to data/embeddings/hyenadna_medium_160k_cds/genes.tsv diff --git a/data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_metadata.json b/data/embeddings/hyenadna_medium_160k_cds/metadata.json similarity index 91% rename from data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_metadata.json rename to data/embeddings/hyenadna_medium_160k_cds/metadata.json index c4e100f15674f23ceb3b6eb82ef4de3f842b9c2f..7c23afe89258c8a48738fea29d6e5cb4492d9fe8 100644 --- a/data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_metadata.json +++ b/data/embeddings/hyenadna_medium_160k_cds/metadata.json @@ -58,5 +58,13 @@ "min_norm": 12.109533309936523, "median_norm": 14.881399154663086, "max_norm": 15.269989967346191 + }, + "embedding_id": "hyenadna_medium_160k_cds", + "former_processing_embedding_id": "hyenadna_medium_160k_cds_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/mahi_all_contexts_mean/README.md b/data/embeddings/mahi_all_contexts_mean/README.md new file mode 100644 index 0000000000000000000000000000000000000000..beb463b1756f3c0cbddbae9eadc6679c90556c4e --- /dev/null +++ b/data/embeddings/mahi_all_contexts_mean/README.md @@ -0,0 +1,68 @@ +# mahi_all_contexts_mean_v1_1 + +Embedding ID: `mahi_all_contexts_mean` + +Former processing ID: `mahi_all_contexts_mean_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18918 genes × 3666 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18918 | 18918 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `entrez_id` | 18918 | 18918 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `mahi_context_count` | 18918 | 1 | 289 | TODO: define this source-specific column during manual curation. | +| `hgnc_symbol` | 18918 | 18918 | TSPAN6
TNMD
DPM1
SCYL3 | HGNC-associated gene symbol. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `embedding_name` | mahi_all_contexts_mean_v1_1 | +| `pooling_strategy` | mean across all non-global tissue/cell contexts per Entrez gene | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_embeddings.npz b/data/embeddings/mahi_all_contexts_mean/embeddings.npz similarity index 100% rename from data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_embeddings.npz rename to data/embeddings/mahi_all_contexts_mean/embeddings.npz diff --git a/data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_genes.tsv b/data/embeddings/mahi_all_contexts_mean/genes.tsv similarity index 100% rename from data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_genes.tsv rename to data/embeddings/mahi_all_contexts_mean/genes.tsv diff --git a/data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_metadata.json b/data/embeddings/mahi_all_contexts_mean/metadata.json similarity index 78% rename from data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_metadata.json rename to data/embeddings/mahi_all_contexts_mean/metadata.json index d34fb5bf92c5475b30158ebb9c7569f97b904060..5532641e63ab7aa1699e8c43ca1f17bc6445edaf 100644 --- a/data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_metadata.json +++ b/data/embeddings/mahi_all_contexts_mean/metadata.json @@ -23,5 +23,13 @@ }, "pooling_strategy": "mean across all non-global tissue/cell contexts per Entrez gene", "master_gene_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv", - "runtime_seconds": 951.99 -} \ No newline at end of file + "runtime_seconds": 951.99, + "embedding_id": "mahi_all_contexts_mean", + "former_processing_embedding_id": "mahi_all_contexts_mean_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/mahi_global/README.md b/data/embeddings/mahi_global/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e69daf9625404668e226f1270208d78b019201ec --- /dev/null +++ b/data/embeddings/mahi_global/README.md @@ -0,0 +1,68 @@ +# mahi_global_v1_1 + +Embedding ID: `mahi_global` + +Former processing ID: `mahi_global_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18918 genes × 3666 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18918 | 18918 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `entrez_id` | 18918 | 18918 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `mahi_context_count` | 18918 | 1 | 1 | TODO: define this source-specific column during manual curation. | +| `hgnc_symbol` | 18918 | 18918 | TSPAN6
TNMD
DPM1
SCYL3 | HGNC-associated gene symbol. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `embedding_name` | mahi_global_v1_1 | +| `pooling_strategy` | global context only | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/mahi_global_v1_1/mahi_global_v1_1_embeddings.npz b/data/embeddings/mahi_global/embeddings.npz similarity index 100% rename from data/embeddings/mahi_global_v1_1/mahi_global_v1_1_embeddings.npz rename to data/embeddings/mahi_global/embeddings.npz diff --git a/data/embeddings/mahi_global_v1_1/mahi_global_v1_1_genes.tsv b/data/embeddings/mahi_global/genes.tsv similarity index 100% rename from data/embeddings/mahi_global_v1_1/mahi_global_v1_1_genes.tsv rename to data/embeddings/mahi_global/genes.tsv diff --git a/data/embeddings/mahi_global_v1_1/mahi_global_v1_1_metadata.json b/data/embeddings/mahi_global/metadata.json similarity index 78% rename from data/embeddings/mahi_global_v1_1/mahi_global_v1_1_metadata.json rename to data/embeddings/mahi_global/metadata.json index d91161f9c8b7fda0ad4aad17a31d79741984b676..0c14d8144198c546579a7283d2020af293d59044 100644 --- a/data/embeddings/mahi_global_v1_1/mahi_global_v1_1_metadata.json +++ b/data/embeddings/mahi_global/metadata.json @@ -23,5 +23,13 @@ }, "pooling_strategy": "global context only", "master_gene_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv", - "runtime_seconds": 951.99 -} \ No newline at end of file + "runtime_seconds": 951.99, + "embedding_id": "mahi_global", + "former_processing_embedding_id": "mahi_global_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/mashup_string/README.md b/data/embeddings/mashup_string/README.md new file mode 100644 index 0000000000000000000000000000000000000000..10b87c67f62e39020d9771195c872be0d6966dba --- /dev/null +++ b/data/embeddings/mashup_string/README.md @@ -0,0 +1,75 @@ +# Mashup-STRING-v1.1 + +Embedding ID: `mashup_string` + +Former processing ID: `mashup_string_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `16331 genes × 800 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 16331 | 16331 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 16331 | 16331 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 16331 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 16327 | 16328 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 16329 | 16282 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 16331 | 16331 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 16331 | 16331 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 16331 | 1 | Mashup-STRING-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 16331 | 3 | gene_symbol
ensembl_protein_id
ensembl_protein_id;gene_symbol | Type of source identifier used by the original embedding. | +| `source_key_count` | 16331 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 16331 | 16331 | TSPAN6
TNMD
DPM1
SCYL3 | Example source keys mapped to this final gene. | +| `mapping_methods` | 16331 | 9 | direct_current_gene_symbol
hgnc_previous_symbol
direct_canonical_protein_id
ensembl_grch37_gtf_protein_id | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | network_ppi | +| `embedding_name` | Mashup-STRING-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/mashup_string_v1_1/mashup_string_v1_1_embeddings.npz b/data/embeddings/mashup_string/embeddings.npz similarity index 100% rename from data/embeddings/mashup_string_v1_1/mashup_string_v1_1_embeddings.npz rename to data/embeddings/mashup_string/embeddings.npz diff --git a/data/embeddings/mashup_string_v1_1/mashup_string_v1_1_genes.tsv b/data/embeddings/mashup_string/genes.tsv similarity index 100% rename from data/embeddings/mashup_string_v1_1/mashup_string_v1_1_genes.tsv rename to data/embeddings/mashup_string/genes.tsv diff --git a/data/embeddings/mashup_string_v1_1/mashup_string_v1_1_metadata.json b/data/embeddings/mashup_string/metadata.json similarity index 83% rename from data/embeddings/mashup_string_v1_1/mashup_string_v1_1_metadata.json rename to data/embeddings/mashup_string/metadata.json index 64e4eb61e2791256ca6811c5dc32d3efd898cd66..36cce1515b331fd83b75f0f204ccff20af44001e 100644 --- a/data/embeddings/mashup_string_v1_1/mashup_string_v1_1_metadata.json +++ b/data/embeddings/mashup_string/metadata.json @@ -29,5 +29,13 @@ "coverage_of_master_rows": 0.8126088470916057, "n_zero_vectors": 0, "output_npz": "/local/home/dtabares/data/processed_embeddings/mashup_string_v1_1/mashup_string_v1_1_embeddings.npz", - "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/mashup_string_v1_1/mashup_string_v1_1_genes.tsv" -} \ No newline at end of file + "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/mashup_string_v1_1/mashup_string_v1_1_genes.tsv", + "embedding_id": "mashup_string", + "former_processing_embedding_id": "mashup_string_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/msa_transformer_depth1_uniprot_human/README.md b/data/embeddings/msa_transformer_depth1_uniprot_human/README.md new file mode 100644 index 0000000000000000000000000000000000000000..df6fc010e44ff5755448b999e65fd6013382a2f5 --- /dev/null +++ b/data/embeddings/msa_transformer_depth1_uniprot_human/README.md @@ -0,0 +1,77 @@ +# MSATransformer-depth1-UniProt-human-v1.1 + +Embedding ID: `msa_transformer_depth1_uniprot_human` + +Former processing ID: `msa_transformer_depth1_uniprot_human_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19472 genes × 768 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `row_index` | 19472 | 19472 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 19472 | 19472 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19472 | 19446 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `hgnc_approved_symbol` | 19336 | 19337 | TSPAN6
TNMD
DPM1
SCYL3 | HGNC-approved gene symbol from the master table. | +| `entrez_id` | 19424 | 19391 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19472 | 19286 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `matched_uniprot_accession` | 19472 | 19267 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession matched during processing. | +| `sequence_length` | 19472 | 2150 | 245
317
260
742 | Input sequence length used for this gene. | +| `n_windows` | 19472 | 11 | 1
2
3
5 | Number of sequence windows used for this gene. | +| `msa_depth` | 19472 | 1 | 1 | Number of sequences used in the MSA input. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein_sequence_msa_transformer_depth1 | +| `embedding_name` | MSATransformer-depth1-UniProt-human-v1.1 | +| `original_model_or_method` | MSA Transformer | +| `embedding_generated_by` | Daniel locally | +| `model_checkpoint` | esm_msa1b_t12_100M_UR50S | +| `pooling_strategy` | mean residue representation from layer 12; long proteins split into non-overlapping windows and length-weighted mean pooled | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_embeddings.npz b/data/embeddings/msa_transformer_depth1_uniprot_human/embeddings.npz similarity index 100% rename from data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_embeddings.npz rename to data/embeddings/msa_transformer_depth1_uniprot_human/embeddings.npz diff --git a/data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_genes.tsv b/data/embeddings/msa_transformer_depth1_uniprot_human/genes.tsv similarity index 100% rename from data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_genes.tsv rename to data/embeddings/msa_transformer_depth1_uniprot_human/genes.tsv diff --git a/data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_metadata.json b/data/embeddings/msa_transformer_depth1_uniprot_human/metadata.json similarity index 80% rename from data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_metadata.json rename to data/embeddings/msa_transformer_depth1_uniprot_human/metadata.json index fa44599d23478dd98ffc388bb616e892c9cdd8aa..482a0e54abc45961a795d63530da45a29a59e38c 100644 --- a/data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_metadata.json +++ b/data/embeddings/msa_transformer_depth1_uniprot_human/metadata.json @@ -29,5 +29,13 @@ "nan_count": 0, "inf_count": 0, "failed_sequence_count": 0, - "runtime_minutes": 7.6259543895721436 -} \ No newline at end of file + "runtime_minutes": 7.6259543895721436, + "embedding_id": "msa_transformer_depth1_uniprot_human", + "former_processing_embedding_id": "msa_transformer_depth1_uniprot_human_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/newt_archs4_256/README.md b/data/embeddings/newt_archs4_256/README.md new file mode 100644 index 0000000000000000000000000000000000000000..707966dd497e72b497ba28357484ef4ac459e652 --- /dev/null +++ b/data/embeddings/newt_archs4_256/README.md @@ -0,0 +1,79 @@ +# NEWT-ARCHS4-256-v1.1 + +Embedding ID: `newt_archs4_256` + +Former processing ID: `newt_archs4_256_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `16265 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 16265 | 16265 | ENSG00000000005
ENSG00000000457
ENSG00000000938
ENSG00000000971 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 16265 | 16265 | TNMD
SCYL3
FGR
CFH | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 16265 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 16265 | 16265 | 64102
57147
2268
3075 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 16260 | 16152 | Q9H2S6
Q8IZE3
P09769
P08603 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 16265 | 16265 | ENST00000373031
ENST00000367771
ENST00000374005
ENST00000367429 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 16265 | 16265 | ENSP00000362122
ENSP00000356745
ENSP00000363117
ENSP00000356399 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 16265 | 1 | NEWT-ARCHS4-256-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 16265 | 1 | Entrez Gene ID from NEWT component embedding CSV | Type of source identifier used by the original embedding. | +| `source_key_count` | 16265 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 16265 | 16265 | 64102
57147
2268
3075 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 16265 | 16265 | 4667
13115
3487
10760 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 16265 | 1 | entrez_id_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT | +| `embedding_name` | NEWT-ARCHS4-256-v1.1 | +| `download_source` | Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318 | +| `original_model_or_method` | NEWT component embedding | +| `embedding_generated_by` | NEWT authors; processed locally into the project standard format | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_embeddings.npz b/data/embeddings/newt_archs4_256/embeddings.npz similarity index 100% rename from data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_embeddings.npz rename to data/embeddings/newt_archs4_256/embeddings.npz diff --git a/data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_genes.tsv b/data/embeddings/newt_archs4_256/genes.tsv similarity index 100% rename from data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_genes.tsv rename to data/embeddings/newt_archs4_256/genes.tsv diff --git a/data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_metadata.json b/data/embeddings/newt_archs4_256/metadata.json similarity index 93% rename from data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_metadata.json rename to data/embeddings/newt_archs4_256/metadata.json index 060c10fb22cb498dbb8081204513bbdba50b615d..006fdf293ed3f8e166c7e10466550a77bedda61b 100644 --- a/data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_metadata.json +++ b/data/embeddings/newt_archs4_256/metadata.json @@ -54,5 +54,13 @@ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/newt_archs4_256_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/newt_archs4_256_v1_1_duplicate_final_mappings.tsv", "excluded_non_entrez_tsv": "/local/home/dtabares/reports/mapping_reports/newt_archs4_256_v1_1_excluded_non_entrez_source_tokens.tsv" + }, + "embedding_id": "newt_archs4_256", + "former_processing_embedding_id": "newt_archs4_256_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/newt_cellnet/README.md b/data/embeddings/newt_cellnet/README.md new file mode 100644 index 0000000000000000000000000000000000000000..08f312b15490d320b71943302e54e7c586a2bafe --- /dev/null +++ b/data/embeddings/newt_cellnet/README.md @@ -0,0 +1,79 @@ +# NEWT-CellNet-v1.1 + +Embedding ID: `newt_cellnet` + +Former processing ID: `newt_cellnet_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `17499 genes × 128 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 17499 | 17499 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 17499 | 17499 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 17499 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 17499 | 17499 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 17498 | 17438 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 17499 | 17499 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 17499 | 17499 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 17499 | 1 | NEWT-CellNet-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 17499 | 1 | Entrez Gene ID from NEWT component embedding CSV | Type of source identifier used by the original embedding. | +| `source_key_count` | 17499 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 17499 | 17499 | 7105
64102
8813
57147 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 17499 | 17499 | 3506
14268
11219
6296 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 17499 | 1 | entrez_id_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT | +| `embedding_name` | NEWT-CellNet-v1.1 | +| `download_source` | Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318 | +| `original_model_or_method` | NEWT component embedding | +| `embedding_generated_by` | NEWT authors; processed locally into the project standard format | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_embeddings.npz b/data/embeddings/newt_cellnet/embeddings.npz similarity index 100% rename from data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_embeddings.npz rename to data/embeddings/newt_cellnet/embeddings.npz diff --git a/data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_genes.tsv b/data/embeddings/newt_cellnet/genes.tsv similarity index 100% rename from data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_genes.tsv rename to data/embeddings/newt_cellnet/genes.tsv diff --git a/data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_metadata.json b/data/embeddings/newt_cellnet/metadata.json similarity index 93% rename from data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_metadata.json rename to data/embeddings/newt_cellnet/metadata.json index cb2741c2bc7788eb96d8ac38daeff1df3ffdb216..956c9171baf2f666b89169623ef488b9fca1fe16 100644 --- a/data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_metadata.json +++ b/data/embeddings/newt_cellnet/metadata.json @@ -54,5 +54,13 @@ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/newt_cellnet_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/newt_cellnet_v1_1_duplicate_final_mappings.tsv", "excluded_non_entrez_tsv": "/local/home/dtabares/reports/mapping_reports/newt_cellnet_v1_1_excluded_non_entrez_source_tokens.tsv" + }, + "embedding_id": "newt_cellnet", + "former_processing_embedding_id": "newt_cellnet_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/newt_go_256/README.md b/data/embeddings/newt_go_256/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e71f1470c458fa95c0d1457863a76dc3bc0aa84b --- /dev/null +++ b/data/embeddings/newt_go_256/README.md @@ -0,0 +1,79 @@ +# NEWT-GO-256-v1.1 + +Embedding ID: `newt_go_256` + +Former processing ID: `newt_go_256_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `13788 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 13788 | 13788 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 13788 | 13788 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 13788 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 13788 | 13788 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 13788 | 13734 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 13788 | 13788 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 13788 | 13788 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 13788 | 1 | NEWT-GO-256-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 13788 | 1 | Entrez Gene ID from NEWT component embedding CSV | Type of source identifier used by the original embedding. | +| `source_key_count` | 13788 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 13788 | 13788 | 7105
64102
8813
57147 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 13788 | 13788 | 2602
14085
2774
3911 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 13788 | 1 | entrez_id_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT | +| `embedding_name` | NEWT-GO-256-v1.1 | +| `download_source` | Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318 | +| `original_model_or_method` | NEWT component embedding | +| `embedding_generated_by` | NEWT authors; processed locally into the project standard format | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_embeddings.npz b/data/embeddings/newt_go_256/embeddings.npz similarity index 100% rename from data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_embeddings.npz rename to data/embeddings/newt_go_256/embeddings.npz diff --git a/data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_genes.tsv b/data/embeddings/newt_go_256/genes.tsv similarity index 100% rename from data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_genes.tsv rename to data/embeddings/newt_go_256/genes.tsv diff --git a/data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_metadata.json b/data/embeddings/newt_go_256/metadata.json similarity index 93% rename from data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_metadata.json rename to data/embeddings/newt_go_256/metadata.json index d3875149dfbce23b534aa8ed02375ef10ffef781..227dd7cc023163c6d5607e20c5d0d004a385ed69 100644 --- a/data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_metadata.json +++ b/data/embeddings/newt_go_256/metadata.json @@ -54,5 +54,13 @@ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/newt_go_256_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/newt_go_256_v1_1_duplicate_final_mappings.tsv", "excluded_non_entrez_tsv": "/local/home/dtabares/reports/mapping_reports/newt_go_256_v1_1_excluded_non_entrez_source_tokens.tsv" + }, + "embedding_id": "newt_go_256", + "former_processing_embedding_id": "newt_go_256_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/newt_go_graph/README.md b/data/embeddings/newt_go_graph/README.md new file mode 100644 index 0000000000000000000000000000000000000000..96a59ac0c349e7daf4385ec69f5451b02a2884e2 --- /dev/null +++ b/data/embeddings/newt_go_graph/README.md @@ -0,0 +1,79 @@ +# NEWT-GO-graph-v1.1 + +Embedding ID: `newt_go_graph` + +Former processing ID: `newt_go_graph_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `13788 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 13788 | 13788 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 13788 | 13788 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 13788 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 13788 | 13788 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 13788 | 13734 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 13788 | 13788 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 13788 | 13788 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 13788 | 1 | NEWT-GO-graph-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 13788 | 1 | Entrez Gene ID from NEWT component embedding CSV | Type of source identifier used by the original embedding. | +| `source_key_count` | 13788 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 13788 | 13788 | 7105
64102
8813
57147 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 13788 | 13788 | 5142
4356
4277
10330 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 13788 | 1 | entrez_id_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT | +| `embedding_name` | NEWT-GO-graph-v1.1 | +| `download_source` | Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318 | +| `original_model_or_method` | NEWT component embedding | +| `embedding_generated_by` | NEWT authors; processed locally into the project standard format | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_embeddings.npz b/data/embeddings/newt_go_graph/embeddings.npz similarity index 100% rename from data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_embeddings.npz rename to data/embeddings/newt_go_graph/embeddings.npz diff --git a/data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_genes.tsv b/data/embeddings/newt_go_graph/genes.tsv similarity index 100% rename from data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_genes.tsv rename to data/embeddings/newt_go_graph/genes.tsv diff --git a/data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_metadata.json b/data/embeddings/newt_go_graph/metadata.json similarity index 93% rename from data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_metadata.json rename to data/embeddings/newt_go_graph/metadata.json index 41a8457377298572f5c42176b9915602210e2d22..0879cc9d405c0c5e1847fb3e7440f4fe8506b96b 100644 --- a/data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_metadata.json +++ b/data/embeddings/newt_go_graph/metadata.json @@ -54,5 +54,13 @@ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/newt_go_graph_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/newt_go_graph_v1_1_duplicate_final_mappings.tsv", "excluded_non_entrez_tsv": "/local/home/dtabares/reports/mapping_reports/newt_go_graph_v1_1_excluded_non_entrez_source_tokens.tsv" + }, + "embedding_id": "newt_go_graph", + "former_processing_embedding_id": "newt_go_graph_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/newt_msigdb_bundle/README.md b/data/embeddings/newt_msigdb_bundle/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0464e54759a77e1ebfc20b5480bca0e34593050e --- /dev/null +++ b/data/embeddings/newt_msigdb_bundle/README.md @@ -0,0 +1,79 @@ +# NEWT-MSigDB-bundle-v1.1 + +Embedding ID: `newt_msigdb_bundle` + +Former processing ID: `newt_msigdb_bundle_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `15423 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 15423 | 15423 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 15423 | 15423 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 15423 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 15423 | 15423 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 15419 | 15388 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 15423 | 15423 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 15423 | 15423 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 15423 | 1 | NEWT-MSigDB-bundle-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 15423 | 1 | Entrez Gene ID from NEWT component embedding CSV | Type of source identifier used by the original embedding. | +| `source_key_count` | 15423 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 15423 | 15423 | 7105
64102
8813
57147 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 15423 | 15423 | 2631
11384
8925
16480 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 15423 | 1 | entrez_id_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT | +| `embedding_name` | NEWT-MSigDB-bundle-v1.1 | +| `download_source` | Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318 | +| `original_model_or_method` | NEWT component embedding | +| `embedding_generated_by` | NEWT authors; processed locally into the project standard format | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_embeddings.npz b/data/embeddings/newt_msigdb_bundle/embeddings.npz similarity index 100% rename from data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_embeddings.npz rename to data/embeddings/newt_msigdb_bundle/embeddings.npz diff --git a/data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_genes.tsv b/data/embeddings/newt_msigdb_bundle/genes.tsv similarity index 100% rename from data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_genes.tsv rename to data/embeddings/newt_msigdb_bundle/genes.tsv diff --git a/data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_metadata.json b/data/embeddings/newt_msigdb_bundle/metadata.json similarity index 93% rename from data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_metadata.json rename to data/embeddings/newt_msigdb_bundle/metadata.json index ac6fec5b8a0dd5e3714cb71982340b5219dac4d3..9538ac1c2f25697b7672bb7c38ccd774a6a3790b 100644 --- a/data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_metadata.json +++ b/data/embeddings/newt_msigdb_bundle/metadata.json @@ -54,5 +54,13 @@ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/newt_msigdb_bundle_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/newt_msigdb_bundle_v1_1_duplicate_final_mappings.tsv", "excluded_non_entrez_tsv": "/local/home/dtabares/reports/mapping_reports/newt_msigdb_bundle_v1_1_excluded_non_entrez_source_tokens.tsv" + }, + "embedding_id": "newt_msigdb_bundle", + "former_processing_embedding_id": "newt_msigdb_bundle_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/node2vec_consensus_ppi/README.md b/data/embeddings/node2vec_consensus_ppi/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c2c9cd02b91d7f8c54a0f2de8e611210d10097a2 --- /dev/null +++ b/data/embeddings/node2vec_consensus_ppi/README.md @@ -0,0 +1,79 @@ +# Node2Vec-consensus-PPI-v1.1 + +Embedding ID: `node2vec_consensus_ppi` + +Former processing ID: `node2vec_consensus_ppi_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18356 genes × 128 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18356 | 18356 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18356 | 18356 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18356 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18356 | 18356 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18354 | 18254 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18356 | 18356 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18356 | 18356 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18356 | 1 | Node2Vec-consensus-PPI-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 18356 | 1 | Entrez Gene ID from consensus PPI graph node | Type of source identifier used by the original embedding. | +| `source_key_count` | 18356 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18356 | 18356 | 7105
64102
8813
57147 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18356 | 18356 | 4958
12201
5934
11631 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18356 | 1 | entrez_id_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | PPI network embedding | +| `embedding_name` | Node2Vec-consensus-PPI-v1.1 | +| `download_source` | ylaboratory/gene-embedding-benchmarks GitHub repository: src/other/preprocess_embedding/consensus.dat | +| `original_model_or_method` | node2vec: scalable feature learning for networks | +| `embedding_generated_by` | Daniel locally from consensus.dat using reproducible random walks and Word2Vec skip-gram | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_embeddings.npz b/data/embeddings/node2vec_consensus_ppi/embeddings.npz similarity index 100% rename from data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_embeddings.npz rename to data/embeddings/node2vec_consensus_ppi/embeddings.npz diff --git a/data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_genes.tsv b/data/embeddings/node2vec_consensus_ppi/genes.tsv similarity index 100% rename from data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_genes.tsv rename to data/embeddings/node2vec_consensus_ppi/genes.tsv diff --git a/data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_metadata.json b/data/embeddings/node2vec_consensus_ppi/metadata.json similarity index 92% rename from data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_metadata.json rename to data/embeddings/node2vec_consensus_ppi/metadata.json index e3715b9435cffa86a21b65836d4fa6e6e408a7ed..7fc8840cb8775992a46a488cdcd56a9969b08ad5 100644 --- a/data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_metadata.json +++ b/data/embeddings/node2vec_consensus_ppi/metadata.json @@ -53,5 +53,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/node2vec_consensus_ppi_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/node2vec_consensus_ppi_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/node2vec_consensus_ppi_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "node2vec_consensus_ppi", + "former_processing_embedding_id": "node2vec_consensus_ppi_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/nt_v2_500m_multispecies_cds/README.md b/data/embeddings/nt_v2_500m_multispecies_cds/README.md new file mode 100644 index 0000000000000000000000000000000000000000..bd3b91ce0775b54c13124b75aba1e3d7f6fca7e6 --- /dev/null +++ b/data/embeddings/nt_v2_500m_multispecies_cds/README.md @@ -0,0 +1,81 @@ +# nt_v2_500m_multispecies_cds_v1_1 + +Embedding ID: `nt_v2_500m_multispecies_cds` + +Former processing ID: `nt_v2_500m_multispecies_cds_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20085 genes × 1024 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `master_row_index` | 20085 | 20085 | 0
1
2
3 | Row index of the corresponding gene in the master gene table. | +| `ensembl_gene_id` | 20085 | 20085 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 20085 | 20068 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `entrez_id` | 19535 | 19511 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 20036 | 19860 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 20085 | 20085 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_transcript_id_versioned` | 20085 | 20085 | ENST00000373020.9
ENST00000373031.5
ENST00000371588.10
ENST00000367771.11 | Versioned canonical transcript identifier. | +| `source_transcript_id_used` | 20085 | 20085 | ENST00000373020.9
ENST00000373031.5
ENST00000371588.10
ENST00000367771.11 | Original transcript identifier selected as model input. | +| `cds_length_nt` | 20085 | 2255 | 738
954
783
2067 | Coding sequence length in nucleotides. | +| `mapping_method` | 20085 | 2 | canonical_transcript_id_versioned_exact
canonical_transcript_id_unversioned_exact | Mapping method used to connect the source record to the master gene table. | +| `nt_v2_token_count_used_for_mean_pooling` | 20085 | 1337 | 123
159
133
347 | TODO: define this source-specific column during manual curation. | +| `effective_chunk_nt` | 20085 | 1 | 6000 | TODO: define this source-specific column during manual curation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | DNA language model / CDS sequence | +| `embedding_name` | nt_v2_500m_multispecies_cds_v1_1 | +| `download_source` | official HuggingFace model repository | +| `original_model_or_method` | Nucleotide Transformer v2 | +| `embedding_generated_by` | Daniel locally by model inference | +| `model_checkpoint` | InstaDeepAI/nucleotide-transformer-v2-500m-multi-species | +| `pooling_strategy` | final hidden layer mean pooling over non-special, non-padding tokens; chunk-weighted by token count for long CDS | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_embeddings.npz b/data/embeddings/nt_v2_500m_multispecies_cds/embeddings.npz similarity index 100% rename from data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_embeddings.npz rename to data/embeddings/nt_v2_500m_multispecies_cds/embeddings.npz diff --git a/data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_genes.tsv b/data/embeddings/nt_v2_500m_multispecies_cds/genes.tsv similarity index 100% rename from data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_genes.tsv rename to data/embeddings/nt_v2_500m_multispecies_cds/genes.tsv diff --git a/data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_metadata.json b/data/embeddings/nt_v2_500m_multispecies_cds/metadata.json similarity index 89% rename from data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_metadata.json rename to data/embeddings/nt_v2_500m_multispecies_cds/metadata.json index a710546f07c1403e06110061a607e335d65fbf63..e63b778c152fed2a4633e8c01ebb15b0d469fcec 100644 --- a/data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_metadata.json +++ b/data/embeddings/nt_v2_500m_multispecies_cds/metadata.json @@ -47,5 +47,13 @@ "The official pretrained NT-v2 checkpoint was used only for inference.", "12 master-table genes lacked canonical CDS sequence in the selected Ensembl CDS FASTA and were excluded from the processed matrix.", "Model was run in fp32 because fp16 caused dtype mismatch in the custom NT model code on boevastudent03." - ] -} \ No newline at end of file + ], + "embedding_id": "nt_v2_500m_multispecies_cds", + "former_processing_embedding_id": "nt_v2_500m_multispecies_cds_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/opa2vec_goa_human/README.md b/data/embeddings/opa2vec_goa_human/README.md new file mode 100644 index 0000000000000000000000000000000000000000..39c17642e5e695082f10cd79e2e2ed57a01eabf3 --- /dev/null +++ b/data/embeddings/opa2vec_goa_human/README.md @@ -0,0 +1,78 @@ +# OPA2Vec-GOA-human-v1.1 + +Embedding ID: `opa2vec_goa_human` + +Former processing ID: `opa2vec_goa_human_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18660 genes × 200 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18660 | 18660 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18660 | 18660 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18660 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18654 | 18655 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18660 | 18660 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18660 | 18660 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18660 | 18660 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18660 | 1 | OPA2Vec-GOA-human-v1.1 | Name of the original source embedding. | +| `source_identifier_type` | 18660 | 1 | UniProt accession | Type of source identifier used by the original embedding. | +| `source_key_count` | 18660 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18660 | 18660 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `mapping_methods` | 18660 | 1 | uniprot_exact | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | Gene Ontology / ontology-annotation semantic embedding | +| `embedding_name` | OPA2Vec-GOA-human-v1.1 | +| `download_source` | Gene Ontology go-basic.obo and GOA human GAF | +| `original_model_or_method` | OPA2Vec | +| `embedding_generated_by` | Daniel locally using an OPA2Vec-derived Python 3 workflow | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_embeddings.npz b/data/embeddings/opa2vec_goa_human/embeddings.npz similarity index 100% rename from data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_embeddings.npz rename to data/embeddings/opa2vec_goa_human/embeddings.npz diff --git a/data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_genes.tsv b/data/embeddings/opa2vec_goa_human/genes.tsv similarity index 100% rename from data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_genes.tsv rename to data/embeddings/opa2vec_goa_human/genes.tsv diff --git a/data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_metadata.json b/data/embeddings/opa2vec_goa_human/metadata.json similarity index 94% rename from data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_metadata.json rename to data/embeddings/opa2vec_goa_human/metadata.json index 1f4a2744a7451fa051879f05d3531e8a878d4a20..74be94b3fd4ff2ec3e40de55d228e438caf24575 100644 --- a/data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_metadata.json +++ b/data/embeddings/opa2vec_goa_human/metadata.json @@ -93,5 +93,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/opa2vec_goa_human_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/opa2vec_goa_human_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/opa2vec_goa_human_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "opa2vec_goa_human", + "former_processing_embedding_id": "opa2vec_goa_human_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/orthrus_base_4track_full_transcript/README.md b/data/embeddings/orthrus_base_4track_full_transcript/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0dd55b96914e7bf7a084b884a6a50540e63e3f5f --- /dev/null +++ b/data/embeddings/orthrus_base_4track_full_transcript/README.md @@ -0,0 +1,69 @@ +# orthrus_base_4track_full_transcript_v1_1 + +Embedding ID: `orthrus_base_4track_full_transcript` + +Former processing ID: `orthrus_base_4track_full_transcript_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19976 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19976 | 19976 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `transcript_id` | 19976 | 19976 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Transcript identifier used for sequence-based processing. | +| `orthrus_sequence_length` | 19976 | 7597 | 3768
1205
1054
6308 | TODO: define this source-specific column during manual curation. | +| `num_chunks` | 19976 | 2 | 1
3 | Number of chunks/windows used when the input was too long. | +| `was_chunked` | 19976 | 2 | False
True | Whether the input sequence was split into chunks. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `embedding_name` | orthrus_base_4track_full_transcript_v1_1 | +| `pooling_strategy` | Orthrus representation() sequence-level embedding. For transcripts longer than max_full_len, length-weighted mean of chunk-level Orthrus embeddings. | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_embeddings.npz b/data/embeddings/orthrus_base_4track_full_transcript/embeddings.npz similarity index 100% rename from data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_embeddings.npz rename to data/embeddings/orthrus_base_4track_full_transcript/embeddings.npz diff --git a/data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_genes.tsv b/data/embeddings/orthrus_base_4track_full_transcript/genes.tsv similarity index 100% rename from data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_genes.tsv rename to data/embeddings/orthrus_base_4track_full_transcript/genes.tsv diff --git a/data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_metadata.json b/data/embeddings/orthrus_base_4track_full_transcript/metadata.json similarity index 82% rename from data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_metadata.json rename to data/embeddings/orthrus_base_4track_full_transcript/metadata.json index ea4730a291d550390673773153236266db590482..74622d7a425ae9f06cb8185f79c01a332e96f936 100644 --- a/data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_metadata.json +++ b/data/embeddings/orthrus_base_4track_full_transcript/metadata.json @@ -26,5 +26,13 @@ "transformers": "4.40.2", "numpy": "1.26.4" }, - "runtime_seconds": 1332.66 -} \ No newline at end of file + "runtime_seconds": 1332.66, + "embedding_id": "orthrus_base_4track_full_transcript", + "former_processing_embedding_id": "orthrus_base_4track_full_transcript_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/README.md b/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e8b696d72625ec23abc8ab81bbaa3bf76457a779 --- /dev/null +++ b/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/README.md @@ -0,0 +1,72 @@ +# PrimeKG-nonPPI-biomedical-context-TFIDF-SVD256-v1.1 + +Embedding ID: `primekg_nonppi_biomedical_context_tfidf_svd256` + +Former processing ID: `primekg_nonppi_biomedical_context_tfidf_svd256_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19216 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `embedding_row` | 19216 | 19216 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 19216 | 19216 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19216 | 19190 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19216 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19216 | 19181 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19209 | 19039 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `source_entrez_id` | 19216 | 19181 | 7105
64102
8813
57147 | Entrez Gene identifier from the original source before final mapping. | +| `source_gene_symbol` | 19216 | 19181 | TSPAN6
TNMD
DPM1
SCYL3 | Original gene symbol from the source before final mapping. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | heterogeneous biomedical knowledge graph / non-PPI gene context | +| `embedding_name` | PrimeKG-nonPPI-biomedical-context-TFIDF-SVD256-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_embeddings.npz b/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/embeddings.npz similarity index 100% rename from data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_embeddings.npz rename to data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/embeddings.npz diff --git a/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_genes.tsv b/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/genes.tsv similarity index 100% rename from data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_genes.tsv rename to data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/genes.tsv diff --git a/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_metadata.json b/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/metadata.json similarity index 91% rename from data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_metadata.json rename to data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/metadata.json index 185d8bf960204b6383d02d6d562f600d8adc154a..d64c3e393d2f657865083fe33d51472d43f230dd 100644 --- a/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_metadata.json +++ b/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/metadata.json @@ -44,5 +44,13 @@ "mapping_report": "/local/home/dtabares/data/processed_embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/mapping_report.txt", "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/unmapped_master_genes.tsv", "unmapped_source_genes": "/local/home/dtabares/reports/mapping_reports/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/unmapped_source_genes.tsv" + }, + "embedding_id": "primekg_nonppi_biomedical_context_tfidf_svd256", + "former_processing_embedding_id": "primekg_nonppi_biomedical_context_tfidf_svd256_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_aac/README.md b/data/embeddings/probe_aac/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ebb941036060faa949b9a7e497144af6510459ea --- /dev/null +++ b/data/embeddings/probe_aac/README.md @@ -0,0 +1,79 @@ +# probe_aac_v1_1 + +Embedding ID: `probe_aac` + +Former processing ID: `probe_aac_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18873 genes × 20 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18873 | 18873 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18873 | 18873 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18873 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18870 | 18871 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18873 | 18873 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18873 | 18873 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18873 | 1 | PROBE_AAC | Name of the original source embedding. | +| `source_identifier_type` | 18873 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18873 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18873 | 18873 | 151
9306
11596
15948 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18873 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_aac_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | AAC | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_aac_v1_1/probe_aac_v1_1_embeddings.npz b/data/embeddings/probe_aac/embeddings.npz similarity index 100% rename from data/embeddings/probe_aac_v1_1/probe_aac_v1_1_embeddings.npz rename to data/embeddings/probe_aac/embeddings.npz diff --git a/data/embeddings/probe_aac_v1_1/probe_aac_v1_1_genes.tsv b/data/embeddings/probe_aac/genes.tsv similarity index 100% rename from data/embeddings/probe_aac_v1_1/probe_aac_v1_1_genes.tsv rename to data/embeddings/probe_aac/genes.tsv diff --git a/data/embeddings/probe_aac_v1_1/probe_aac_v1_1_metadata.json b/data/embeddings/probe_aac/metadata.json similarity index 92% rename from data/embeddings/probe_aac_v1_1/probe_aac_v1_1_metadata.json rename to data/embeddings/probe_aac/metadata.json index dd2ec284afc3173d4b662fd4bb8790c3ea13f9cf..9688a6703d312d25dc92a2ffe4445ba1b984adc3 100644 --- a/data/embeddings/probe_aac_v1_1/probe_aac_v1_1_metadata.json +++ b/data/embeddings/probe_aac/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_aac_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_aac_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_aac_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_aac", + "former_processing_embedding_id": "probe_aac_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_albert/README.md b/data/embeddings/probe_albert/README.md new file mode 100644 index 0000000000000000000000000000000000000000..7ffbdf886683f42a25aadac179f320ed436e3a5f --- /dev/null +++ b/data/embeddings/probe_albert/README.md @@ -0,0 +1,79 @@ +# probe_albert_v1_1 + +Embedding ID: `probe_albert` + +Former processing ID: `probe_albert_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18453 genes × 4096 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18453 | 18453 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18453 | 18453 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18453 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18450 | 18451 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18453 | 18453 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18453 | 18453 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18453 | 1 | PROBE_ALBERT | Name of the original source embedding. | +| `source_identifier_type` | 18453 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18453 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18453 | 18453 | 17995
17597
4682
12159 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18453 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_albert_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | ALBERT | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_albert_v1_1/probe_albert_v1_1_embeddings.npz b/data/embeddings/probe_albert/embeddings.npz similarity index 100% rename from data/embeddings/probe_albert_v1_1/probe_albert_v1_1_embeddings.npz rename to data/embeddings/probe_albert/embeddings.npz diff --git a/data/embeddings/probe_albert_v1_1/probe_albert_v1_1_genes.tsv b/data/embeddings/probe_albert/genes.tsv similarity index 100% rename from data/embeddings/probe_albert_v1_1/probe_albert_v1_1_genes.tsv rename to data/embeddings/probe_albert/genes.tsv diff --git a/data/embeddings/probe_albert_v1_1/probe_albert_v1_1_metadata.json b/data/embeddings/probe_albert/metadata.json similarity index 92% rename from data/embeddings/probe_albert_v1_1/probe_albert_v1_1_metadata.json rename to data/embeddings/probe_albert/metadata.json index c8717e72b079bf5f418880372a74ef68448bcadc..387df20bbabc7821dd1155c1f25a831362090f9b 100644 --- a/data/embeddings/probe_albert_v1_1/probe_albert_v1_1_metadata.json +++ b/data/embeddings/probe_albert/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_albert_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_albert_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_albert_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_albert", + "former_processing_embedding_id": "probe_albert_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_apaac/README.md b/data/embeddings/probe_apaac/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0bb4410691f86373b4b24c929e5903164e629f41 --- /dev/null +++ b/data/embeddings/probe_apaac/README.md @@ -0,0 +1,79 @@ +# probe_apaac_v1_1 + +Embedding ID: `probe_apaac` + +Former processing ID: `probe_apaac_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18871 genes × 80 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18871 | 18871 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18871 | 18871 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18871 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18868 | 18869 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18871 | 18871 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18871 | 18871 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18871 | 18871 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18871 | 1 | PROBE_APAAC | Name of the original source embedding. | +| `source_identifier_type` | 18871 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18871 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18871 | 18871 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18871 | 18871 | 151
9278
11562
15832 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18871 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_apaac_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | APAAC | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_embeddings.npz b/data/embeddings/probe_apaac/embeddings.npz similarity index 100% rename from data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_embeddings.npz rename to data/embeddings/probe_apaac/embeddings.npz diff --git a/data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_genes.tsv b/data/embeddings/probe_apaac/genes.tsv similarity index 100% rename from data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_genes.tsv rename to data/embeddings/probe_apaac/genes.tsv diff --git a/data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_metadata.json b/data/embeddings/probe_apaac/metadata.json similarity index 92% rename from data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_metadata.json rename to data/embeddings/probe_apaac/metadata.json index 4d433c20dab95f28c440317355f179ea81ec8a0b..204ead82cfeecac268ae467389944b09dba614db 100644 --- a/data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_metadata.json +++ b/data/embeddings/probe_apaac/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_apaac_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_apaac_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_apaac_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_apaac", + "former_processing_embedding_id": "probe_apaac_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_bert_bfd/README.md b/data/embeddings/probe_bert_bfd/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a663f21b5008188531a2c1cad887ff8a92e60149 --- /dev/null +++ b/data/embeddings/probe_bert_bfd/README.md @@ -0,0 +1,79 @@ +# probe_bert_bfd_v1_1 + +Embedding ID: `probe_bert_bfd` + +Former processing ID: `probe_bert_bfd_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18453 genes × 1024 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18453 | 18453 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18453 | 18453 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18453 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18450 | 18451 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18453 | 18453 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18453 | 18453 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18453 | 1 | PROBE_BERT-BFD | Name of the original source embedding. | +| `source_identifier_type` | 18453 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18453 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18453 | 18453 | 17995
17597
4682
12159 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18453 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_bert_bfd_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | BERT-BFD | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_embeddings.npz b/data/embeddings/probe_bert_bfd/embeddings.npz similarity index 100% rename from data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_embeddings.npz rename to data/embeddings/probe_bert_bfd/embeddings.npz diff --git a/data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_genes.tsv b/data/embeddings/probe_bert_bfd/genes.tsv similarity index 100% rename from data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_genes.tsv rename to data/embeddings/probe_bert_bfd/genes.tsv diff --git a/data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_metadata.json b/data/embeddings/probe_bert_bfd/metadata.json similarity index 92% rename from data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_metadata.json rename to data/embeddings/probe_bert_bfd/metadata.json index b8ee9520b48b2256bbc135149abb1d574388fdc7..d72d770b7a0e6d4c05dec9130acff89785dab837 100644 --- a/data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_metadata.json +++ b/data/embeddings/probe_bert_bfd/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_bert_bfd_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_bert_bfd_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_bert_bfd_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_bert_bfd", + "former_processing_embedding_id": "probe_bert_bfd_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_bert_pfam/README.md b/data/embeddings/probe_bert_pfam/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6069533f02d0a588a5556d06d84936786d70c35b --- /dev/null +++ b/data/embeddings/probe_bert_pfam/README.md @@ -0,0 +1,79 @@ +# probe_bert_pfam_v1_1 + +Embedding ID: `probe_bert_pfam` + +Former processing ID: `probe_bert_pfam_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18461 genes × 768 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18461 | 18461 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18461 | 18461 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18461 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18458 | 18459 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18461 | 18461 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18461 | 18461 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18461 | 18461 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18461 | 1 | PROBE_BERT-PFAM | Name of the original source embedding. | +| `source_identifier_type` | 18461 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18461 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18461 | 18461 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18461 | 18461 | 2547
18524
11316
17195 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18461 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_bert_pfam_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | BERT-PFAM | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_embeddings.npz b/data/embeddings/probe_bert_pfam/embeddings.npz similarity index 100% rename from data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_embeddings.npz rename to data/embeddings/probe_bert_pfam/embeddings.npz diff --git a/data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_genes.tsv b/data/embeddings/probe_bert_pfam/genes.tsv similarity index 100% rename from data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_genes.tsv rename to data/embeddings/probe_bert_pfam/genes.tsv diff --git a/data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_metadata.json b/data/embeddings/probe_bert_pfam/metadata.json similarity index 92% rename from data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_metadata.json rename to data/embeddings/probe_bert_pfam/metadata.json index ec7de34e7e56fd0e26a088a852283bdfff6e0723..562df73b429f70e0c41d8b0e83f85704f8146be3 100644 --- a/data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_metadata.json +++ b/data/embeddings/probe_bert_pfam/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_bert_pfam_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_bert_pfam_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_bert_pfam_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_bert_pfam", + "former_processing_embedding_id": "probe_bert_pfam_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_blast/README.md b/data/embeddings/probe_blast/README.md new file mode 100644 index 0000000000000000000000000000000000000000..a1199b8ee3b781135670149bf17428b2fbdcced7 --- /dev/null +++ b/data/embeddings/probe_blast/README.md @@ -0,0 +1,79 @@ +# probe_blast_v1_1 + +Embedding ID: `probe_blast` + +Former processing ID: `probe_blast_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18873 genes × 20421 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18873 | 18873 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18873 | 18873 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18873 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18870 | 18871 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18873 | 18873 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18873 | 18873 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18873 | 1 | PROBE_BLAST | Name of the original source embedding. | +| `source_identifier_type` | 18873 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18873 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18873 | 18873 | 151
9306
11596
15948 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18873 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_blast_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | BLAST | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_blast_v1_1/probe_blast_v1_1_embeddings.npz b/data/embeddings/probe_blast/embeddings.npz similarity index 100% rename from data/embeddings/probe_blast_v1_1/probe_blast_v1_1_embeddings.npz rename to data/embeddings/probe_blast/embeddings.npz diff --git a/data/embeddings/probe_blast_v1_1/probe_blast_v1_1_genes.tsv b/data/embeddings/probe_blast/genes.tsv similarity index 100% rename from data/embeddings/probe_blast_v1_1/probe_blast_v1_1_genes.tsv rename to data/embeddings/probe_blast/genes.tsv diff --git a/data/embeddings/probe_blast_v1_1/probe_blast_v1_1_metadata.json b/data/embeddings/probe_blast/metadata.json similarity index 92% rename from data/embeddings/probe_blast_v1_1/probe_blast_v1_1_metadata.json rename to data/embeddings/probe_blast/metadata.json index af76112aa9332a3baba3a66a90f30797bff1adb9..e3f821629ed8934dcfb569ed95dd39db4bdbe873 100644 --- a/data/embeddings/probe_blast_v1_1/probe_blast_v1_1_metadata.json +++ b/data/embeddings/probe_blast/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_blast_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_blast_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_blast_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_blast", + "former_processing_embedding_id": "probe_blast_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_cpc_prot/README.md b/data/embeddings/probe_cpc_prot/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f6769fefe25d0f2c620ab6d30a706954ae23270 --- /dev/null +++ b/data/embeddings/probe_cpc_prot/README.md @@ -0,0 +1,79 @@ +# probe_cpc_prot_v1_1 + +Embedding ID: `probe_cpc_prot` + +Former processing ID: `probe_cpc_prot_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18453 genes × 512 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18453 | 18453 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18453 | 18453 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18453 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18450 | 18451 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18453 | 18453 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18453 | 18453 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18453 | 1 | PROBE_CPC-PROT | Name of the original source embedding. | +| `source_identifier_type` | 18453 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18453 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18453 | 18453 | 17995
17597
4682
12159 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18453 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_cpc_prot_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | CPC-PROT | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_embeddings.npz b/data/embeddings/probe_cpc_prot/embeddings.npz similarity index 100% rename from data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_embeddings.npz rename to data/embeddings/probe_cpc_prot/embeddings.npz diff --git a/data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_genes.tsv b/data/embeddings/probe_cpc_prot/genes.tsv similarity index 100% rename from data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_genes.tsv rename to data/embeddings/probe_cpc_prot/genes.tsv diff --git a/data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_metadata.json b/data/embeddings/probe_cpc_prot/metadata.json similarity index 92% rename from data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_metadata.json rename to data/embeddings/probe_cpc_prot/metadata.json index 764b272cd6dc4743052f5397f116c93679814825..cc004975c8320678376272d971b66ee5e5442b03 100644 --- a/data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_metadata.json +++ b/data/embeddings/probe_cpc_prot/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_cpc_prot_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_cpc_prot_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_cpc_prot_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_cpc_prot", + "former_processing_embedding_id": "probe_cpc_prot_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_esmb1/README.md b/data/embeddings/probe_esmb1/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2f9fec6e929c69587adc36690a9a9e515cfdc8e9 --- /dev/null +++ b/data/embeddings/probe_esmb1/README.md @@ -0,0 +1,79 @@ +# probe_esmb1_v1_1 + +Embedding ID: `probe_esmb1` + +Former processing ID: `probe_esmb1_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18453 genes × 1280 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18453 | 18453 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18453 | 18453 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18453 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18450 | 18451 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18453 | 18453 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18453 | 18453 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18453 | 1 | PROBE_ESMB1 | Name of the original source embedding. | +| `source_identifier_type` | 18453 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18453 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18453 | 18453 | 17995
17597
4682
12159 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18453 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_esmb1_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | ESMB1 | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_embeddings.npz b/data/embeddings/probe_esmb1/embeddings.npz similarity index 100% rename from data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_embeddings.npz rename to data/embeddings/probe_esmb1/embeddings.npz diff --git a/data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_genes.tsv b/data/embeddings/probe_esmb1/genes.tsv similarity index 100% rename from data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_genes.tsv rename to data/embeddings/probe_esmb1/genes.tsv diff --git a/data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_metadata.json b/data/embeddings/probe_esmb1/metadata.json similarity index 92% rename from data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_metadata.json rename to data/embeddings/probe_esmb1/metadata.json index 2556fd06111d424dc28696ff4069e0296cef27d1..beb2d4727f86fa82310177d55a3ea23443acc575 100644 --- a/data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_metadata.json +++ b/data/embeddings/probe_esmb1/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_esmb1_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_esmb1_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_esmb1_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_esmb1", + "former_processing_embedding_id": "probe_esmb1_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_gene2vec_uniprot/README.md b/data/embeddings/probe_gene2vec_uniprot/README.md new file mode 100644 index 0000000000000000000000000000000000000000..796eb1696c9a14bca60917dfa81a400dee32a9a7 --- /dev/null +++ b/data/embeddings/probe_gene2vec_uniprot/README.md @@ -0,0 +1,79 @@ +# probe_gene2vec_uniprot_v1_1 + +Embedding ID: `probe_gene2vec_uniprot` + +Former processing ID: `probe_gene2vec_uniprot_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `17564 genes × 200 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 17564 | 17564 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 17564 | 17564 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 17564 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 17563 | 17564 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 17564 | 17564 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 17564 | 17564 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 17564 | 17564 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 17564 | 1 | PROBE_GENE2VEC | Name of the original source embedding. | +| `source_identifier_type` | 17564 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 17564 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 17564 | 17564 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 17564 | 17564 | 123
8286
10390
14334 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 17564 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_gene2vec_uniprot_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | GENE2VEC | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_embeddings.npz b/data/embeddings/probe_gene2vec_uniprot/embeddings.npz similarity index 100% rename from data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_embeddings.npz rename to data/embeddings/probe_gene2vec_uniprot/embeddings.npz diff --git a/data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_genes.tsv b/data/embeddings/probe_gene2vec_uniprot/genes.tsv similarity index 100% rename from data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_genes.tsv rename to data/embeddings/probe_gene2vec_uniprot/genes.tsv diff --git a/data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_metadata.json b/data/embeddings/probe_gene2vec_uniprot/metadata.json similarity index 92% rename from data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_metadata.json rename to data/embeddings/probe_gene2vec_uniprot/metadata.json index a53c97022691a226d244ce376f40b1bec8dfaa26..2def65446392c4f060b27a4053b6ed421e342ee9 100644 --- a/data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_metadata.json +++ b/data/embeddings/probe_gene2vec_uniprot/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_gene2vec_uniprot_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_gene2vec_uniprot_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_gene2vec_uniprot_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_gene2vec_uniprot", + "former_processing_embedding_id": "probe_gene2vec_uniprot_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_hmmer/README.md b/data/embeddings/probe_hmmer/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b428ee277358ffbc3055a4e0bfd555eb23326b93 --- /dev/null +++ b/data/embeddings/probe_hmmer/README.md @@ -0,0 +1,79 @@ +# probe_hmmer_v1_1 + +Embedding ID: `probe_hmmer` + +Former processing ID: `probe_hmmer_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18873 genes × 20421 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18873 | 18873 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18873 | 18873 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18873 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18870 | 18871 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18873 | 18873 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18873 | 18873 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18873 | 1 | PROBE_HMMER | Name of the original source embedding. | +| `source_identifier_type` | 18873 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18873 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18873 | 18873 | 151
9306
11596
15948 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18873 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_hmmer_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | HMMER | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_embeddings.npz b/data/embeddings/probe_hmmer/embeddings.npz similarity index 100% rename from data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_embeddings.npz rename to data/embeddings/probe_hmmer/embeddings.npz diff --git a/data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_genes.tsv b/data/embeddings/probe_hmmer/genes.tsv similarity index 100% rename from data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_genes.tsv rename to data/embeddings/probe_hmmer/genes.tsv diff --git a/data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_metadata.json b/data/embeddings/probe_hmmer/metadata.json similarity index 92% rename from data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_metadata.json rename to data/embeddings/probe_hmmer/metadata.json index 600494418ec7c8298fe46c7af7ce5394ac1961ff..33b4bcafd013d8531ad7bc030d8c9c552fc1fc2c 100644 --- a/data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_metadata.json +++ b/data/embeddings/probe_hmmer/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_hmmer_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_hmmer_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_hmmer_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_hmmer", + "former_processing_embedding_id": "probe_hmmer_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_ksep/README.md b/data/embeddings/probe_ksep/README.md new file mode 100644 index 0000000000000000000000000000000000000000..b20a4137b53edd4d2741ef2cf9909cb2c22f2a6f --- /dev/null +++ b/data/embeddings/probe_ksep/README.md @@ -0,0 +1,79 @@ +# probe_ksep_v1_1 + +Embedding ID: `probe_ksep` + +Former processing ID: `probe_ksep_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18801 genes × 400 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18801 | 18801 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18801 | 18801 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18801 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18798 | 18799 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18801 | 18801 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18801 | 18801 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18801 | 18801 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18801 | 1 | PROBE_KSEP | Name of the original source embedding. | +| `source_identifier_type` | 18801 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18801 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18801 | 18801 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18801 | 18801 | 150
9225
11509
15765 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18801 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_ksep_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | KSEP | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_embeddings.npz b/data/embeddings/probe_ksep/embeddings.npz similarity index 100% rename from data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_embeddings.npz rename to data/embeddings/probe_ksep/embeddings.npz diff --git a/data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_genes.tsv b/data/embeddings/probe_ksep/genes.tsv similarity index 100% rename from data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_genes.tsv rename to data/embeddings/probe_ksep/genes.tsv diff --git a/data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_metadata.json b/data/embeddings/probe_ksep/metadata.json similarity index 92% rename from data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_metadata.json rename to data/embeddings/probe_ksep/metadata.json index a56bc8ef332ede4e6f1f47bb1e48d85c565ecda4..7b4c36f946ba19bb6ee44f486847026180f789be 100644 --- a/data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_metadata.json +++ b/data/embeddings/probe_ksep/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_ksep_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_ksep_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_ksep_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_ksep", + "former_processing_embedding_id": "probe_ksep_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_learned_vec/README.md b/data/embeddings/probe_learned_vec/README.md new file mode 100644 index 0000000000000000000000000000000000000000..8d7005432df43f2d56b1aa9e7593267f4bfe1b32 --- /dev/null +++ b/data/embeddings/probe_learned_vec/README.md @@ -0,0 +1,79 @@ +# probe_learned_vec_v1_1 + +Embedding ID: `probe_learned_vec` + +Former processing ID: `probe_learned_vec_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18873 genes × 64 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18873 | 18873 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18873 | 18873 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18873 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18870 | 18871 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18873 | 18873 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18873 | 18873 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18873 | 1 | PROBE_LEARNED-VEC | Name of the original source embedding. | +| `source_identifier_type` | 18873 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18873 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18873 | 18873 | 18377
17975
4793
12441 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18873 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_learned_vec_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | LEARNED-VEC | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_embeddings.npz b/data/embeddings/probe_learned_vec/embeddings.npz similarity index 100% rename from data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_embeddings.npz rename to data/embeddings/probe_learned_vec/embeddings.npz diff --git a/data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_genes.tsv b/data/embeddings/probe_learned_vec/genes.tsv similarity index 100% rename from data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_genes.tsv rename to data/embeddings/probe_learned_vec/genes.tsv diff --git a/data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_metadata.json b/data/embeddings/probe_learned_vec/metadata.json similarity index 92% rename from data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_metadata.json rename to data/embeddings/probe_learned_vec/metadata.json index fe1f8b69ca09846d508ff32417d6c321c7096c30..f2041ad9f6792bd3ec9793a2890833cd7da8b2e6 100644 --- a/data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_metadata.json +++ b/data/embeddings/probe_learned_vec/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_learned_vec_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_learned_vec_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_learned_vec_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_learned_vec", + "former_processing_embedding_id": "probe_learned_vec_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_mut2vec/README.md b/data/embeddings/probe_mut2vec/README.md new file mode 100644 index 0000000000000000000000000000000000000000..5dcc4cd6686bc689e0b20a0f3546e4d266e9a87b --- /dev/null +++ b/data/embeddings/probe_mut2vec/README.md @@ -0,0 +1,79 @@ +# probe_mut2vec_v1_1 + +Embedding ID: `probe_mut2vec` + +Former processing ID: `probe_mut2vec_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `17445 genes × 300 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 17445 | 17445 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 17445 | 17445 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 17445 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 17444 | 17445 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 17445 | 17445 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 17445 | 17445 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 17445 | 17445 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 17445 | 1 | PROBE_MUT2VEC | Name of the original source embedding. | +| `source_identifier_type` | 17445 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 17445 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 17445 | 17445 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 17445 | 17445 | 16661
12935
13923
9616 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 17445 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_mut2vec_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | MUT2VEC | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_embeddings.npz b/data/embeddings/probe_mut2vec/embeddings.npz similarity index 100% rename from data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_embeddings.npz rename to data/embeddings/probe_mut2vec/embeddings.npz diff --git a/data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_genes.tsv b/data/embeddings/probe_mut2vec/genes.tsv similarity index 100% rename from data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_genes.tsv rename to data/embeddings/probe_mut2vec/genes.tsv diff --git a/data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_metadata.json b/data/embeddings/probe_mut2vec/metadata.json similarity index 92% rename from data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_metadata.json rename to data/embeddings/probe_mut2vec/metadata.json index 1b98833324ad6b029b7335d5f877359638cce9f1..80b1469549306f906b88704af8a08723e2d3d80e 100644 --- a/data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_metadata.json +++ b/data/embeddings/probe_mut2vec/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_mut2vec_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_mut2vec_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_mut2vec_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_mut2vec", + "former_processing_embedding_id": "probe_mut2vec_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_pfam/README.md b/data/embeddings/probe_pfam/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4b3c91b8e09f39a393ebda2ddde79cf27cc1961c --- /dev/null +++ b/data/embeddings/probe_pfam/README.md @@ -0,0 +1,79 @@ +# probe_pfam_v1_1 + +Embedding ID: `probe_pfam` + +Former processing ID: `probe_pfam_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18873 genes × 6227 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18873 | 18873 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18873 | 18873 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18873 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18870 | 18871 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18873 | 18873 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18873 | 18873 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18873 | 1 | PROBE_PFAM | Name of the original source embedding. | +| `source_identifier_type` | 18873 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18873 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18873 | 18873 | 151
9304
11594
15946 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18873 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_pfam_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | PFAM | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_embeddings.npz b/data/embeddings/probe_pfam/embeddings.npz similarity index 100% rename from data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_embeddings.npz rename to data/embeddings/probe_pfam/embeddings.npz diff --git a/data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_genes.tsv b/data/embeddings/probe_pfam/genes.tsv similarity index 100% rename from data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_genes.tsv rename to data/embeddings/probe_pfam/genes.tsv diff --git a/data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_metadata.json b/data/embeddings/probe_pfam/metadata.json similarity index 92% rename from data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_metadata.json rename to data/embeddings/probe_pfam/metadata.json index 47cc386657ea245d4e3f92958de3816af11fb6fb..1f7ecce75a9308caec52d38ba18d5fc7cef4c54e 100644 --- a/data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_metadata.json +++ b/data/embeddings/probe_pfam/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_pfam_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_pfam_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_pfam_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_pfam", + "former_processing_embedding_id": "probe_pfam_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_protvec/README.md b/data/embeddings/probe_protvec/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0236b5ac587156f07b94d24699ae0d77f11c5976 --- /dev/null +++ b/data/embeddings/probe_protvec/README.md @@ -0,0 +1,79 @@ +# probe_protvec_v1_1 + +Embedding ID: `probe_protvec` + +Former processing ID: `probe_protvec_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18873 genes × 100 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18873 | 18873 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18873 | 18873 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18873 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18870 | 18871 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18873 | 18873 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18873 | 18873 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18873 | 1 | PROBE_PROTVEC | Name of the original source embedding. | +| `source_identifier_type` | 18873 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18873 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18873 | 18873 | 18377
17975
4793
12441 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18873 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_protvec_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | PROTVEC | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_embeddings.npz b/data/embeddings/probe_protvec/embeddings.npz similarity index 100% rename from data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_embeddings.npz rename to data/embeddings/probe_protvec/embeddings.npz diff --git a/data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_genes.tsv b/data/embeddings/probe_protvec/genes.tsv similarity index 100% rename from data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_genes.tsv rename to data/embeddings/probe_protvec/genes.tsv diff --git a/data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_metadata.json b/data/embeddings/probe_protvec/metadata.json similarity index 92% rename from data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_metadata.json rename to data/embeddings/probe_protvec/metadata.json index 888297df9a55325cf394bab4e968a5e6194e4950..160a25f73b500bcca2497a357e7eb94cd7a2ef7e 100644 --- a/data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_metadata.json +++ b/data/embeddings/probe_protvec/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_protvec_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_protvec_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_protvec_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_protvec", + "former_processing_embedding_id": "probe_protvec_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_seqvec/README.md b/data/embeddings/probe_seqvec/README.md new file mode 100644 index 0000000000000000000000000000000000000000..6de6a901f11bab8a5d1fd2fff4439cf81c14c7ab --- /dev/null +++ b/data/embeddings/probe_seqvec/README.md @@ -0,0 +1,79 @@ +# probe_seqvec_v1_1 + +Embedding ID: `probe_seqvec` + +Former processing ID: `probe_seqvec_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18873 genes × 1024 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18873 | 18873 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18873 | 18873 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18873 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18870 | 18871 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18873 | 18873 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18873 | 18873 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18873 | 1 | PROBE_SEQVEC | Name of the original source embedding. | +| `source_identifier_type` | 18873 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18873 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18873 | 18873 | 18377
17975
4793
12441 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18873 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_seqvec_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | SEQVEC | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_embeddings.npz b/data/embeddings/probe_seqvec/embeddings.npz similarity index 100% rename from data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_embeddings.npz rename to data/embeddings/probe_seqvec/embeddings.npz diff --git a/data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_genes.tsv b/data/embeddings/probe_seqvec/genes.tsv similarity index 100% rename from data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_genes.tsv rename to data/embeddings/probe_seqvec/genes.tsv diff --git a/data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_metadata.json b/data/embeddings/probe_seqvec/metadata.json similarity index 92% rename from data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_metadata.json rename to data/embeddings/probe_seqvec/metadata.json index a668723c8356b6a054e4d642dfb459e926ad85f9..fe94177c566eb170de13ae8d86d1fe77cca23c00 100644 --- a/data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_metadata.json +++ b/data/embeddings/probe_seqvec/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_seqvec_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_seqvec_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_seqvec_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_seqvec", + "former_processing_embedding_id": "probe_seqvec_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_t5/README.md b/data/embeddings/probe_t5/README.md new file mode 100644 index 0000000000000000000000000000000000000000..09d00c86ae5896173a00478a393a3e16d3f4c147 --- /dev/null +++ b/data/embeddings/probe_t5/README.md @@ -0,0 +1,79 @@ +# probe_t5_v1_1 + +Embedding ID: `probe_t5` + +Former processing ID: `probe_t5_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18453 genes × 1024 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18453 | 18453 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18453 | 18453 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18453 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18450 | 18451 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18453 | 18453 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18453 | 18453 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18453 | 1 | PROBE_T5 | Name of the original source embedding. | +| `source_identifier_type` | 18453 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18453 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18453 | 18453 | 17995
17597
4682
12159 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18453 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_t5_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | T5 | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_t5_v1_1/probe_t5_v1_1_embeddings.npz b/data/embeddings/probe_t5/embeddings.npz similarity index 100% rename from data/embeddings/probe_t5_v1_1/probe_t5_v1_1_embeddings.npz rename to data/embeddings/probe_t5/embeddings.npz diff --git a/data/embeddings/probe_t5_v1_1/probe_t5_v1_1_genes.tsv b/data/embeddings/probe_t5/genes.tsv similarity index 100% rename from data/embeddings/probe_t5_v1_1/probe_t5_v1_1_genes.tsv rename to data/embeddings/probe_t5/genes.tsv diff --git a/data/embeddings/probe_t5_v1_1/probe_t5_v1_1_metadata.json b/data/embeddings/probe_t5/metadata.json similarity index 92% rename from data/embeddings/probe_t5_v1_1/probe_t5_v1_1_metadata.json rename to data/embeddings/probe_t5/metadata.json index 375980798d022e302867b5f2d39189efb94e0da2..ed1a5bd805d4e7113d31f5dcfb0502494d6e714b 100644 --- a/data/embeddings/probe_t5_v1_1/probe_t5_v1_1_metadata.json +++ b/data/embeddings/probe_t5/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_t5_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_t5_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_t5_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_t5", + "former_processing_embedding_id": "probe_t5_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_tcga_embedding/README.md b/data/embeddings/probe_tcga_embedding/README.md new file mode 100644 index 0000000000000000000000000000000000000000..1ff94642be605ed005624d8d686edad2d9d27c5a --- /dev/null +++ b/data/embeddings/probe_tcga_embedding/README.md @@ -0,0 +1,79 @@ +# probe_tcga_embedding_v1_1 + +Embedding ID: `probe_tcga_embedding` + +Former processing ID: `probe_tcga_embedding_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18221 genes × 50 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18221 | 18221 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18221 | 18221 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18221 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18220 | 18221 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18221 | 18221 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18221 | 18221 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18221 | 18221 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18221 | 1 | PROBE_TCGA-EMBEDDING | Name of the original source embedding. | +| `source_identifier_type` | 18221 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18221 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18221 | 18221 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18221 | 18221 | 123
8466
10602
14770 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18221 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_tcga_embedding_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | TCGA-EMBEDDING | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_embeddings.npz b/data/embeddings/probe_tcga_embedding/embeddings.npz similarity index 100% rename from data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_embeddings.npz rename to data/embeddings/probe_tcga_embedding/embeddings.npz diff --git a/data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_genes.tsv b/data/embeddings/probe_tcga_embedding/genes.tsv similarity index 100% rename from data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_genes.tsv rename to data/embeddings/probe_tcga_embedding/genes.tsv diff --git a/data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_metadata.json b/data/embeddings/probe_tcga_embedding/metadata.json similarity index 92% rename from data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_metadata.json rename to data/embeddings/probe_tcga_embedding/metadata.json index db0e6412b1d6b380899abb3945cf36b7185d860c..1ea5853a6a2c85173a289cafa14ef15729d37cd8 100644 --- a/data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_metadata.json +++ b/data/embeddings/probe_tcga_embedding/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_tcga_embedding_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_tcga_embedding_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_tcga_embedding_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_tcga_embedding", + "former_processing_embedding_id": "probe_tcga_embedding_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_unirep/README.md b/data/embeddings/probe_unirep/README.md new file mode 100644 index 0000000000000000000000000000000000000000..95bab6265ed83edce3dec79ba6b2099d6298a553 --- /dev/null +++ b/data/embeddings/probe_unirep/README.md @@ -0,0 +1,79 @@ +# probe_unirep_v1_1 + +Embedding ID: `probe_unirep` + +Former processing ID: `probe_unirep_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18873 genes × 5700 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18873 | 18873 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18873 | 18873 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18873 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18870 | 18871 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18873 | 18873 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18873 | 18873 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18873 | 1 | PROBE_UNIREP | Name of the original source embedding. | +| `source_identifier_type` | 18873 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18873 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18873 | 18873 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18873 | 18873 | 7655
8683
16444
3518 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18873 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_unirep_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | UNIREP | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_embeddings.npz b/data/embeddings/probe_unirep/embeddings.npz similarity index 100% rename from data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_embeddings.npz rename to data/embeddings/probe_unirep/embeddings.npz diff --git a/data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_genes.tsv b/data/embeddings/probe_unirep/genes.tsv similarity index 100% rename from data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_genes.tsv rename to data/embeddings/probe_unirep/genes.tsv diff --git a/data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_metadata.json b/data/embeddings/probe_unirep/metadata.json similarity index 92% rename from data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_metadata.json rename to data/embeddings/probe_unirep/metadata.json index a54e2868bd427ea4b8b2247095f94ddbb558bf41..0a4aa0224f734a8e66249385de4a36b6a290cc71 100644 --- a/data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_metadata.json +++ b/data/embeddings/probe_unirep/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_unirep_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_unirep_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_unirep_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_unirep", + "former_processing_embedding_id": "probe_unirep_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/probe_xlnet/README.md b/data/embeddings/probe_xlnet/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ee8db00e93312e67c6dd3f5b8dcccb5a200fde43 --- /dev/null +++ b/data/embeddings/probe_xlnet/README.md @@ -0,0 +1,79 @@ +# probe_xlnet_v1_1 + +Embedding ID: `probe_xlnet` + +Former processing ID: `probe_xlnet_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `18453 genes × 1024 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 18453 | 18453 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 18453 | 18453 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 18453 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 18450 | 18451 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 18453 | 18453 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 18453 | 18453 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 18453 | 1 | PROBE_XLNET | Name of the original source embedding. | +| `source_identifier_type` | 18453 | 1 | UniProt accession from PROBE Entry column | Type of source identifier used by the original embedding. | +| `source_key_count` | 18453 | 3 | 1
2
4 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 18453 | 18453 | O43657
Q9H2S6
O60762
Q8IZE3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 18453 | 18453 | 17995
17597
4682
12159 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 18453 | 1 | uniprot_accession_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | protein sequence / protein representation from PROBE human UniProt archive | +| `embedding_name` | probe_xlnet_v1_1 | +| `download_source` | Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository | +| `original_model_or_method` | XLNET | +| `embedding_generated_by` | PROBE authors / original PROBE human representation archive | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_embeddings.npz b/data/embeddings/probe_xlnet/embeddings.npz similarity index 100% rename from data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_embeddings.npz rename to data/embeddings/probe_xlnet/embeddings.npz diff --git a/data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_genes.tsv b/data/embeddings/probe_xlnet/genes.tsv similarity index 100% rename from data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_genes.tsv rename to data/embeddings/probe_xlnet/genes.tsv diff --git a/data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_metadata.json b/data/embeddings/probe_xlnet/metadata.json similarity index 92% rename from data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_metadata.json rename to data/embeddings/probe_xlnet/metadata.json index 89536ac93bf6841861b05d375397cc5048766476..88385b49f7bc57323ee16e6767cd1d0902ae50c9 100644 --- a/data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_metadata.json +++ b/data/embeddings/probe_xlnet/metadata.json @@ -47,5 +47,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/probe_xlnet_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/probe_xlnet_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/probe_xlnet_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "probe_xlnet", + "former_processing_embedding_id": "probe_xlnet_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/pubmedbert_gene_protein_summary/README.md b/data/embeddings/pubmedbert_gene_protein_summary/README.md new file mode 100644 index 0000000000000000000000000000000000000000..fa4e9130894f5983be3d9c3cfa5030996e9828cc --- /dev/null +++ b/data/embeddings/pubmedbert_gene_protein_summary/README.md @@ -0,0 +1,75 @@ +# PubMedBERT-GeneProteinSummary-v1.1 + +Embedding ID: `pubmedbert_gene_protein_summary` + +Former processing ID: `pubmedbert_gene_protein_summary_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19168 genes × 768 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `row_index` | 19168 | 19168 | 0
1
2
3 | Zero-based row index in this embedding matrix and genes.tsv file. | +| `ensembl_gene_id` | 19168 | 19168 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19168 | 19143 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `hgnc_approved_symbol` | 19082 | 19083 | TSPAN6
TNMD
DPM1
SCYL3 | HGNC-approved gene symbol from the master table. | +| `entrez_id` | 19118 | 19092 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19153 | 18980 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `summary_key` | 19168 | 19082 | TSPAN6
TNMD
DPM1
SCYL3 | Identifier used to retrieve the gene/protein text summary. | +| `summary_mapping_source` | 19168 | 3 | gene_symbol
map_symbols_all_fallback
hgnc_approved_symbol | Mapping route used to select the gene/protein text summary. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | literature_text_gene_summary | +| `embedding_name` | PubMedBERT-GeneProteinSummary-v1.1 | +| `download_source` | GenePT Zenodo v2 summaries | +| `embedding_generated_by` | Daniel locally | +| `model_checkpoint` | microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext | +| `pooling_strategy` | final hidden layer mean pooling over valid non-special, non-padding tokens | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_embeddings.npz b/data/embeddings/pubmedbert_gene_protein_summary/embeddings.npz similarity index 100% rename from data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_embeddings.npz rename to data/embeddings/pubmedbert_gene_protein_summary/embeddings.npz diff --git a/data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_genes.tsv b/data/embeddings/pubmedbert_gene_protein_summary/genes.tsv similarity index 100% rename from data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_genes.tsv rename to data/embeddings/pubmedbert_gene_protein_summary/genes.tsv diff --git a/data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_metadata.json b/data/embeddings/pubmedbert_gene_protein_summary/metadata.json similarity index 75% rename from data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_metadata.json rename to data/embeddings/pubmedbert_gene_protein_summary/metadata.json index aef428a5eaefb18b472dcca9abd9036e840e9a06..6c99d805de002c1e3aa9aa5c73be60507ed6e0bf 100644 --- a/data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_metadata.json +++ b/data/embeddings/pubmedbert_gene_protein_summary/metadata.json @@ -23,5 +23,13 @@ "zero_vectors": 0, "nan_count": 0, "inf_count": 0, - "runtime_minutes": 2.671167786916097 -} \ No newline at end of file + "runtime_minutes": 2.671167786916097, + "embedding_id": "pubmedbert_gene_protein_summary", + "former_processing_embedding_id": "pubmedbert_gene_protein_summary_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/rnafm_full_transcript/README.md b/data/embeddings/rnafm_full_transcript/README.md new file mode 100644 index 0000000000000000000000000000000000000000..0abdaa910d452e8d6bcf9ce6dfb8c572f1f62078 --- /dev/null +++ b/data/embeddings/rnafm_full_transcript/README.md @@ -0,0 +1,69 @@ +# rnafm_full_transcript_v1_1 + +Embedding ID: `rnafm_full_transcript` + +Former processing ID: `rnafm_full_transcript_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19976 genes × 640 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19976 | 19976 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `rnafm_sequence_length` | 19976 | 7597 | 3768
1205
1054
6308 | TODO: define this source-specific column during manual curation. | +| `num_chunks` | 19976 | 37 | 5
2
8
3 | Number of chunks/windows used when the input was too long. | +| `transcript_id` | 19976 | 19976 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Transcript identifier used for sequence-based processing. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `embedding_name` | rnafm_full_transcript_v1_1 | +| `model_checkpoint` | /local/home/dtabares/data/raw_embeddings/rnafm_full_transcript_v1_1/hf_model/RNA-FM_pretrained.pth | +| `pooling_strategy` | Final layer 12 mean token pooling over real RNA tokens only; BOS/EOS excluded; long transcripts split into chunks and chunk means length-weighted | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_embeddings.npz b/data/embeddings/rnafm_full_transcript/embeddings.npz similarity index 100% rename from data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_embeddings.npz rename to data/embeddings/rnafm_full_transcript/embeddings.npz diff --git a/data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_genes.tsv b/data/embeddings/rnafm_full_transcript/genes.tsv similarity index 100% rename from data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_genes.tsv rename to data/embeddings/rnafm_full_transcript/genes.tsv diff --git a/data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_metadata.json b/data/embeddings/rnafm_full_transcript/metadata.json similarity index 83% rename from data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_metadata.json rename to data/embeddings/rnafm_full_transcript/metadata.json index 4634fd953dd3bf04819f5b30f526f37b2e673353..3de76fe4b39569d4a38f42f0f8c7228a0552f86c 100644 --- a/data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_metadata.json +++ b/data/embeddings/rnafm_full_transcript/metadata.json @@ -25,5 +25,13 @@ "min_chunk_len": 0, "batch_size": 1, "master_gene_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv", - "runtime_seconds": 3351.46 -} \ No newline at end of file + "runtime_seconds": 3351.46, + "embedding_id": "rnafm_full_transcript", + "former_processing_embedding_id": "rnafm_full_transcript_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/scfoundation_gene_pos/README.md b/data/embeddings/scfoundation_gene_pos/README.md new file mode 100644 index 0000000000000000000000000000000000000000..eeb313ed1146209e40aaa24fad22bd5cca0f3158 --- /dev/null +++ b/data/embeddings/scfoundation_gene_pos/README.md @@ -0,0 +1,80 @@ +# scFoundation-gene-pos-v1.1 + +Embedding ID: `scfoundation_gene_pos` + +Former processing ID: `scfoundation_gene_pos_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19062 genes × 768 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19062 | 19062 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19062 | 19062 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19062 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19062 | 19061 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19062 | 18930 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 19062 | 19062 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 19062 | 19062 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 19062 | 1 | official_scFoundation_gene_checkpoint_pos_emb | Name of the original source embedding. | +| `source_identifier_type` | 19062 | 1 | scFoundation OS_scRNA_gene_index gene symbol | Type of source identifier used by the original embedding. | +| `source_key_count` | 19062 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 19062 | 19062 | TSPAN6
TNMD
DPM1
SCYL3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 19062 | 19062 | 17543
17159
4516
14443 | Example source row indices mapped to this final gene. | +| `source_checkpoint_indices_examples` | 19062 | 19062 | 17543
17159
4516
14443 | TODO: define this source-specific column during manual curation. | +| `mapping_methods` | 19062 | 5 | symbol_current_or_hgnc_approved_exact_unique
symbol_hgnc_previous_exact_unique
symbol_current_or_hgnc_approved_exact_unique;symbol_hgnc_alias_exact_unique
symbol_current_or_hgnc_approved_exact_unique;symbol_hgnc_previous_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | single_cell_expression_foundation_model / learned gene identity-position embedding | +| `embedding_name` | scFoundation-gene-pos-v1.1 | +| `download_source` | Official scFoundation GitHub repository and official SharePoint checkpoint linked from the repository | +| `original_model_or_method` | scFoundation / xTrimoGene | +| `embedding_generated_by` | Extracted locally from official scFoundation gene sub-checkpoint model.pos_emb.weight | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_embeddings.npz b/data/embeddings/scfoundation_gene_pos/embeddings.npz similarity index 100% rename from data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_embeddings.npz rename to data/embeddings/scfoundation_gene_pos/embeddings.npz diff --git a/data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_genes.tsv b/data/embeddings/scfoundation_gene_pos/genes.tsv similarity index 100% rename from data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_genes.tsv rename to data/embeddings/scfoundation_gene_pos/genes.tsv diff --git a/data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_metadata.json b/data/embeddings/scfoundation_gene_pos/metadata.json similarity index 95% rename from data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_metadata.json rename to data/embeddings/scfoundation_gene_pos/metadata.json index 95621be0293ffd1555ee86b6ddbd1b98710c4743..363806e8649a3990fd5337a74a619c6dfff9bcd2 100644 --- a/data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_metadata.json +++ b/data/embeddings/scfoundation_gene_pos/metadata.json @@ -78,5 +78,13 @@ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/scfoundation_gene_pos_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/scfoundation_gene_pos_v1_1_duplicate_final_mappings.tsv", "excluded_special_checkpoint_rows_tsv": "/local/home/dtabares/reports/mapping_reports/scfoundation_gene_pos_v1_1_excluded_special_checkpoint_rows.tsv" + }, + "embedding_id": "scfoundation_gene_pos", + "former_processing_embedding_id": "scfoundation_gene_pos_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/scgpt_pancancer/README.md b/data/embeddings/scgpt_pancancer/README.md new file mode 100644 index 0000000000000000000000000000000000000000..c2daaa81f68561b487a4e929e6f2670b259cf226 --- /dev/null +++ b/data/embeddings/scgpt_pancancer/README.md @@ -0,0 +1,70 @@ +# scgpt_pancancer_v1_1 + +Embedding ID: `scgpt_pancancer` + +Former processing ID: `scgpt_pancancer_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19071 genes × 512 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19071 | 19071 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19071 | 19071 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `source_gene_symbol` | 19071 | 19071 | TSPAN6
TNMD
DPM1
SCYL3 | Original gene symbol from the source before final mapping. | +| `source_vocab_id` | 19071 | 19071 | 34494
33849
7983
30793 | Original vocabulary/token ID from the source model. | +| `symbol_mapping_source` | 19071 | 2 | primary
alias;primary | Symbol mapping route used to connect a source symbol to the master gene table. | +| `entrez_id` | 19071 | 19070 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19066 | 18916 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `embedding_name` | scgpt_pancancer_v1_1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_embeddings.npz b/data/embeddings/scgpt_pancancer/embeddings.npz similarity index 100% rename from data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_embeddings.npz rename to data/embeddings/scgpt_pancancer/embeddings.npz diff --git a/data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_genes.tsv b/data/embeddings/scgpt_pancancer/genes.tsv similarity index 100% rename from data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_genes.tsv rename to data/embeddings/scgpt_pancancer/genes.tsv diff --git a/data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_metadata.json b/data/embeddings/scgpt_pancancer/metadata.json similarity index 90% rename from data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_metadata.json rename to data/embeddings/scgpt_pancancer/metadata.json index 57c7a3017223a43c52a4f0abbffb5c286f6c140a..84dd323f3189edcdcc06ad3d0206bf78a7396007 100644 --- a/data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_metadata.json +++ b/data/embeddings/scgpt_pancancer/metadata.json @@ -59,5 +59,13 @@ "d_hid": 512 }, "master_gene_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv", - "runtime_seconds": 4.92 -} \ No newline at end of file + "runtime_seconds": 4.92, + "embedding_id": "scgpt_pancancer", + "former_processing_embedding_id": "scgpt_pancancer_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/scgpt_whole_human/README.md b/data/embeddings/scgpt_whole_human/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4e08aa88934d9c15e27a0be1e95477a5550695d5 --- /dev/null +++ b/data/embeddings/scgpt_whole_human/README.md @@ -0,0 +1,78 @@ +# scGPT-whole-human-v1.1 + +Embedding ID: `scgpt_whole_human` + +Former processing ID: `scgpt_whole_human_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19280 genes × 512 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19280 | 19280 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19280 | 19280 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19280 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19280 | 19279 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19275 | 19124 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 19280 | 19280 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 19280 | 19280 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 19280 | 1 | official_scGPT_whole_human | Name of the original source embedding. | +| `source_identifier_type` | 19280 | 1 | scGPT vocab gene symbol | Type of source identifier used by the original embedding. | +| `source_key_count` | 19280 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 19280 | 19280 | TSPAN6
TNMD
DPM1
SCYL3 | Example source keys mapped to this final gene. | +| `source_token_ids_examples` | 19280 | 19280 | 34494
33849
7983
30793 | Example source token IDs mapped to this final gene. | +| `mapping_methods` | 19280 | 4 | symbol_current_or_hgnc_approved_exact_unique
symbol_hgnc_previous_exact_unique
symbol_current_or_hgnc_approved_exact_unique;symbol_hgnc_alias_exact_unique
symbol_current_or_hgnc_approved_exact_unique;symbol_hgnc_previous_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | single_cell_expression_foundation_model | +| `embedding_name` | scGPT-whole-human-v1.1 | +| `download_source` | Official scGPT whole-human Google Drive checkpoint linked from the scGPT project | +| `embedding_generated_by` | Extracted locally from official scGPT whole-human model input gene-token embedding layer | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_embeddings.npz b/data/embeddings/scgpt_whole_human/embeddings.npz similarity index 100% rename from data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_embeddings.npz rename to data/embeddings/scgpt_whole_human/embeddings.npz diff --git a/data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_genes.tsv b/data/embeddings/scgpt_whole_human/genes.tsv similarity index 100% rename from data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_genes.tsv rename to data/embeddings/scgpt_whole_human/genes.tsv diff --git a/data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_metadata.json b/data/embeddings/scgpt_whole_human/metadata.json similarity index 92% rename from data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_metadata.json rename to data/embeddings/scgpt_whole_human/metadata.json index 690b4727cb226062c82245efc7accf058d505a16..d3a5a01f9758f05657711191d354a2a29e7014a2 100644 --- a/data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_metadata.json +++ b/data/embeddings/scgpt_whole_human/metadata.json @@ -46,5 +46,13 @@ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/scgpt_whole_human_v1_1_unmapped_source_identifiers.tsv", "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/scgpt_whole_human_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/scgpt_whole_human_v1_1_duplicate_final_mappings.tsv" + }, + "embedding_id": "scgpt_whole_human", + "former_processing_embedding_id": "scgpt_whole_human_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/scprint_medium_v1_5_gene/README.md b/data/embeddings/scprint_medium_v1_5_gene/README.md new file mode 100644 index 0000000000000000000000000000000000000000..49042f802285145f2494d2b54e1a866164e37f26 --- /dev/null +++ b/data/embeddings/scprint_medium_v1_5_gene/README.md @@ -0,0 +1,79 @@ +# scPRINT-medium-v1.5-gene-v1.1 + +Embedding ID: `scprint_medium_v1_5_gene` + +Former processing ID: `scprint_medium_v1_5_gene_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19995 genes × 256 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19995 | 19995 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19995 | 19970 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19995 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19477 | 19445 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19949 | 19765 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 19995 | 19995 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 19995 | 19995 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 19995 | 1 | official_scPRINT_medium_v1_5_gene_encoder_embedding | Name of the original source embedding. | +| `source_identifier_type` | 19995 | 1 | scPRINT checkpoint hyper_parameters genes Ensembl gene ID | Type of source identifier used by the original embedding. | +| `source_key_count` | 19995 | 1 | 1 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 19995 | 19995 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 19995 | 19995 | 0
1
2
3 | Example source row indices mapped to this final gene. | +| `mapping_methods` | 19995 | 1 | ensembl_gene_id_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | single_cell_expression_foundation_model / learned gene-token embedding | +| `embedding_name` | scPRINT-medium-v1.5-gene-v1.1 | +| `download_source` | Official scPRINT HuggingFace repository jkobject/scPRINT | +| `original_model_or_method` | scPRINT | +| `embedding_generated_by` | Extracted locally from official scPRINT medium-v1.5 checkpoint | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_embeddings.npz b/data/embeddings/scprint_medium_v1_5_gene/embeddings.npz similarity index 100% rename from data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_embeddings.npz rename to data/embeddings/scprint_medium_v1_5_gene/embeddings.npz diff --git a/data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_genes.tsv b/data/embeddings/scprint_medium_v1_5_gene/genes.tsv similarity index 100% rename from data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_genes.tsv rename to data/embeddings/scprint_medium_v1_5_gene/genes.tsv diff --git a/data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_metadata.json b/data/embeddings/scprint_medium_v1_5_gene/metadata.json similarity index 93% rename from data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_metadata.json rename to data/embeddings/scprint_medium_v1_5_gene/metadata.json index a09c4a98fb513698b4aeab749413ca1cbf7c6c9e..1970c180190f35dbedc47b767d92df1d7f54ab86 100644 --- a/data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_metadata.json +++ b/data/embeddings/scprint_medium_v1_5_gene/metadata.json @@ -66,5 +66,13 @@ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/scprint_medium_v1_5_gene_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/scprint_medium_v1_5_gene_v1_1_duplicate_final_mappings.tsv", "excluded_nonhuman_or_non_ENSG_rows_tsv": "/local/home/dtabares/reports/mapping_reports/scprint_medium_v1_5_gene_v1_1_excluded_nonhuman_or_non_ensg_rows.tsv" + }, + "embedding_id": "scprint_medium_v1_5_gene", + "former_processing_embedding_id": "scprint_medium_v1_5_gene_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/embeddings/spliceai_canonical_splice_site_profile/README.md b/data/embeddings/spliceai_canonical_splice_site_profile/README.md new file mode 100644 index 0000000000000000000000000000000000000000..75495bdc131cf709edfdf880c1df52bb3b934392 --- /dev/null +++ b/data/embeddings/spliceai_canonical_splice_site_profile/README.md @@ -0,0 +1,79 @@ +# SpliceAI-canonical-splice-site-profile-zscore-v1.1 + +Embedding ID: `spliceai_canonical_splice_site_profile` + +Former processing ID: `spliceai_canonical_splice_site_profile_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `20097 genes × 606 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 20097 | 20097 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 20097 | 20070 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `entrez_id` | 19545 | 19511 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 20045 | 19860 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `gene_type` | 20097 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `chromosome` | 20097 | 25 | chrX
chr20
chr1
chr6 | Chromosome for the gene or modeled interval. | +| `strand` | 20097 | 2 | -
+ | Genomic strand. | +| `canonical_transcript_id` | 20097 | 20097 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_transcript_id_versioned` | 20097 | 20097 | ENST00000373020.9
ENST00000373031.5
ENST00000371588.10
ENST00000367771.11 | Versioned canonical transcript identifier. | +| `assembly` | 20097 | 1 | GRCh38 | Genome assembly used for coordinates. | +| `gencode_release` | 20097 | 1 | 49 | GENCODE release used for annotation. | +| `spliceai_site_status` | 20097 | 2 | ok
single_exon_or_no_junctions | TODO: define this source-specific column during manual curation. | +| `n_splice_sites` | 20097 | 107 | 14
12
16
24 | TODO: define this source-specific column during manual curation. | +| `n_donor_sites` | 20097 | 107 | 7
6
8
12 | TODO: define this source-specific column during manual curation. | +| `n_acceptor_sites` | 20097 | 107 | 7
6
8
12 | TODO: define this source-specific column during manual curation. | + +## Basic metadata + +| Field | Value | +|---|---| +| `status` | processed_qc_passed | +| `modality` | splicing sequence model / canonical transcript splice-site profiles | +| `embedding_name` | SpliceAI-canonical-splice-site-profile-zscore-v1.1 | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_embeddings.npz b/data/embeddings/spliceai_canonical_splice_site_profile/embeddings.npz similarity index 100% rename from data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_embeddings.npz rename to data/embeddings/spliceai_canonical_splice_site_profile/embeddings.npz diff --git a/data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_genes.tsv b/data/embeddings/spliceai_canonical_splice_site_profile/genes.tsv similarity index 100% rename from data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_genes.tsv rename to data/embeddings/spliceai_canonical_splice_site_profile/genes.tsv diff --git a/data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_metadata.json b/data/embeddings/spliceai_canonical_splice_site_profile/metadata.json similarity index 85% rename from data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_metadata.json rename to data/embeddings/spliceai_canonical_splice_site_profile/metadata.json index e9e50d73df2c3f06ef8d0818a336516090c19e94..833c58975b980944fd1c9730f5e11a3590c4b9ec 100644 --- a/data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_metadata.json +++ b/data/embeddings/spliceai_canonical_splice_site_profile/metadata.json @@ -35,5 +35,13 @@ }, "final_embedding": "dimension-wise z-scored mean SpliceAI acceptor/donor profiles across canonical splice sites", "genes_without_canonical_splice_sites": 1151, - "created_at": "2026-07-10T01:30:35" -} \ No newline at end of file + "created_at": "2026-07-10T01:30:35", + "embedding_id": "spliceai_canonical_splice_site_profile", + "former_processing_embedding_id": "spliceai_canonical_splice_site_profile_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" + } +} diff --git a/data/embeddings/uce_33l/README.md b/data/embeddings/uce_33l/README.md new file mode 100644 index 0000000000000000000000000000000000000000..9e37d843c2794b5be9f12b7d382a1b598d977e8d --- /dev/null +++ b/data/embeddings/uce_33l/README.md @@ -0,0 +1,81 @@ +# UCE-33L-v1.1 + +Embedding ID: `uce_33l` + +Former processing ID: `uce_33l_v1_1` + +## Public files + +- `embeddings.npz`: compressed NumPy archive containing the embedding matrix under key `embeddings`. +- `genes.tsv`: row metadata table. Row order matches the embedding matrix. +- `metadata.json`: machine-readable provenance and processing metadata. +- `README.md`: this per-embedding documentation file. + +## Matrix + +- Shape: `19082 genes × 1280 dimensions` +- Dtype: `float32` +- NPZ array key: `embeddings` + +The row order is authoritative: + +```text +embeddings.npz['embeddings'][i, :] corresponds to genes.tsv row i +``` + +## Loading example + +```python +import numpy as np +import pandas as pd + +with np.load('embeddings.npz') as z: + X = z['embeddings'] + +genes = pd.read_csv('genes.tsv', sep='\t') +assert X.shape[0] == len(genes) +``` + +## genes.tsv columns + +| Column | Non-empty values | Unique values | Example values | Meaning | +|---|---:|---:|---|---| +| `ensembl_gene_id` | 19082 | 19082 | ENSG00000000003
ENSG00000000005
ENSG00000000419
ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. | +| `gene_symbol` | 19082 | 19082 | TSPAN6
TNMD
DPM1
SCYL3 | Gene symbol associated with the Ensembl gene ID. | +| `gene_type` | 19082 | 1 | protein_coding | Gene biotype, usually protein_coding. | +| `entrez_id` | 19082 | 19081 | 7105
64102
8813
57147 | NCBI Entrez Gene identifier, when available. | +| `uniprot_id` | 19075 | 18933 | O43657
Q9H2S6
O60762
Q8IZE3 | UniProt accession or identifier, when available. | +| `canonical_transcript_id` | 19082 | 19082 | ENST00000373020
ENST00000373031
ENST00000371588
ENST00000367771 | Canonical transcript identifier recorded or used for this gene. | +| `canonical_protein_id` | 19082 | 19082 | ENSP00000362111
ENSP00000362122
ENSP00000360644
ENSP00000356745 | Canonical protein identifier recorded or used for this gene. | +| `source_embedding` | 19082 | 1 | official_UCE_33L_8ep_1024t_1280 | Name of the original source embedding. | +| `source_identifier_type` | 19082 | 1 | UCE species_chrom human gene symbol | Type of source identifier used by the original embedding. | +| `source_key_count` | 19082 | 2 | 1
2 | Number of source keys mapped or collapsed to this final gene. | +| `source_keys_examples` | 19082 | 19082 | TSPAN6
TNMD
DPM1
SCYL3 | Example source keys mapped to this final gene. | +| `source_row_indices_examples` | 19082 | 19082 | 18962
18961
12000
750 | Example source row indices mapped to this final gene. | +| `source_chromosomes_examples` | 19082 | 32 | X
20
1
6 | TODO: define this source-specific column during manual curation. | +| `source_starts_examples` | 19082 | 19078 | 100632063
100584936
50936148
169853074 | TODO: define this source-specific column during manual curation. | +| `mapping_methods` | 19082 | 3 | symbol_current_or_hgnc_approved_exact_unique
symbol_hgnc_previous_exact_unique
symbol_current_or_hgnc_approved_exact_unique;symbol_hgnc_previous_exact_unique | Mapping methods used to connect source records to the master gene table. | + +## Basic metadata + +| Field | Value | +|---|---| +| `modality` | single_cell_expression_foundation_model / protein-informed gene-token embedding | +| `embedding_name` | UCE-33L-v1.1 | +| `download_source` | Official UCE GitHub repository plus official Figshare model files | +| `original_model_or_method` | Universal Cell Embeddings (UCE) | +| `embedding_generated_by` | Extracted locally from official UCE all_tokens and 33-layer checkpoint gene encoder | + +## Detailed processing notes + +TODO: manually paste and curate the processing explanation for this embedding from the central project README. + +Please include, when relevant: + +- original source/model/checkpoint; +- whether the embedding was downloaded or generated locally; +- input data used; +- identifier mapping and filtering decisions; +- pooling, aggregation, PCA, or normalization method; +- coverage and important caveats; +- license or redistribution notes. diff --git a/data/embeddings/uce_33l_v1_1/uce_33l_v1_1_embeddings.npz b/data/embeddings/uce_33l/embeddings.npz similarity index 100% rename from data/embeddings/uce_33l_v1_1/uce_33l_v1_1_embeddings.npz rename to data/embeddings/uce_33l/embeddings.npz diff --git a/data/embeddings/uce_33l_v1_1/uce_33l_v1_1_genes.tsv b/data/embeddings/uce_33l/genes.tsv similarity index 100% rename from data/embeddings/uce_33l_v1_1/uce_33l_v1_1_genes.tsv rename to data/embeddings/uce_33l/genes.tsv diff --git a/data/embeddings/uce_33l_v1_1/uce_33l_v1_1_metadata.json b/data/embeddings/uce_33l/metadata.json similarity index 95% rename from data/embeddings/uce_33l_v1_1/uce_33l_v1_1_metadata.json rename to data/embeddings/uce_33l/metadata.json index 212a8483b5e3d9c358f7a13aa3ddbf0f993b320d..1265266c1d1223bd5786ca0d1d0d80a3ce9d5e84 100644 --- a/data/embeddings/uce_33l_v1_1/uce_33l_v1_1_metadata.json +++ b/data/embeddings/uce_33l/metadata.json @@ -85,5 +85,13 @@ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/uce_33l_v1_1_ambiguous_source_identifiers.tsv", "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/uce_33l_v1_1_duplicate_final_mappings.tsv", "excluded_unembedded_human_rows_tsv": "/local/home/dtabares/reports/mapping_reports/uce_33l_v1_1_excluded_unembedded_human_source_rows.tsv" + }, + "embedding_id": "uce_33l", + "former_processing_embedding_id": "uce_33l_v1_1", + "public_file_structure": { + "embeddings": "embeddings.npz", + "genes": "genes.tsv", + "metadata": "metadata.json", + "readme": "README.md" } -} \ No newline at end of file +} diff --git a/data/manifest.csv b/data/manifest.csv index 65d270b5b40a6502fc2d5de3d734a18798977629..e675dce2090edeb0e783bb4ff56107f639cfaade 100644 --- a/data/manifest.csv +++ b/data/manifest.csv @@ -1,69 +1,69 @@ -embedding_id,n_genes,dimension,dtype,npz_array_key,embeddings_npz,genes_tsv,metadata_json,status,modality,embedding_name,download_source,original_model_or_method,embedding_generated_by,model_checkpoint,pooling_strategy,license,file_size_bytes -alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1,20084,256,float32,embeddings,data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_embeddings.npz,data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_genes.tsv,data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_metadata.json,processed_qc_passed,,AlphaGenome-TSS1Mb-regulatoryTrackSummary-PCA256-v1.1,,,,,,,19270344 -basenji2_tss_131kb_human_central3_v1_1,20097,5313,float32,embeddings,data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_embeddings.npz,data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_genes.tsv,data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_metadata.json,processed_qc_passed,regulatory DNA / Basenji2 TSS sequence model,Basenji2-TSS-131kb-human-central3-log1p-zscore-v1.1,,,,,,,389641898 -biobert_gene_protein_summary_v1_1,19168,768,float32,embeddings,data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_embeddings.npz,data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_genes.tsv,data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_metadata.json,,literature_text_gene_summary,BioBERT-GeneProteinSummary-v1.1,GenePT Zenodo v2 summaries,,Daniel locally,dmis-lab/biobert-base-cased-v1.2,"final hidden layer mean pooling over valid non-special, non-padding tokens",,54606212 -bioconceptvec_cbow_v1_1,17040,100,float32,embeddings,data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_embeddings.npz,data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_genes.tsv,data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_metadata.json,,biomedical_literature_text,BioConceptVec-CBOW-v1.1,,,,,,,6339061 -bioconceptvec_fasttext_v1_1,17040,100,float32,embeddings,data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_embeddings.npz,data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_genes.tsv,data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_metadata.json,,biomedical_literature_text,BioConceptVec-fastText-v1.1,,,,,,,6254929 -bioconceptvec_glove_v1_1,17040,100,float32,embeddings,data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_embeddings.npz,data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_genes.tsv,data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_metadata.json,,biomedical_literature_text,BioConceptVec-GloVe-v1.1,,,,,,,6350146 -bioconceptvec_skipgram_v1_1,17040,100,float32,embeddings,data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_embeddings.npz,data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_genes.tsv,data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_metadata.json,,biomedical_literature_text,BioConceptVec-SkipGram-v1.1,,,,,,,6335039 -cmap_gse106127_cgs_mean_landmark978_zscore_v1_1,4070,978,float32,embeddings,data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_embeddings.npz,data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_genes.tsv,data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_metadata.json,processed_qc_passed,perturbational transcriptomics / L1000 consensus genetic perturbation signatures,CMap-GSE106127-CGS-mean-landmark978-zscore-v1.1,,,,,,,14768959 -deepnf_string_v12_ppmi_svd_fusion_v1_1,19071,256,float32,embeddings,data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_embeddings.npz,data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_genes.tsv,data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_metadata.json,,STRING network multimodal sparse PPMI/SVD fusion,deepNF-STRING-v12-PPMI-SVD-fusion-v1.1,"STRING v12 human protein.links.detailed, protein.info, and protein.aliases files",deepNF: deep network fusion for protein function prediction,Daniel locally from STRING v12 channel networks,,,,18225284 -deepsea_tss_1kb_epigenomic_probs_v1_1,20097,919,float32,embeddings,data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_embeddings.npz,data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_genes.tsv,data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_metadata.json,processed_qc_passed,regulatory DNA / promoter-TSS sequence model,DeepSEA-TSS-1kb-epigenomic-probabilities-zscore-v1.1,,,,,,,66380222 -depmap_crispr_gene_effect_pca256_v1_1,18562,256,float32,embeddings,data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_embeddings.npz,data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_genes.tsv,data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_metadata.json,processed_qc_passed,functional genomics / CRISPR dependency,DepMap-CRISPRGeneEffect-PCA256-v1.1,,,,,,,17729467 -dnabert2_117m_cds_v1_1,20085,768,float32,embeddings,data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_embeddings.npz,data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_genes.tsv,data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_metadata.json,processed_qc_passed,,dnabert2_117m_cds_v1_1,,,,,"Mean pooling over final-layer non-special token embeddings. Long CDS sequences split into nucleotide chunks, then length-weighted mean of chunk-level embeddings.",,57063502 -enformer_tss_393kb_human_central3_v1_1,20097,5313,float32,embeddings,data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_embeddings.npz,data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_genes.tsv,data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_metadata.json,processed_qc_passed,regulatory DNA / long-range TSS sequence model,Enformer-TSS-393kb-human-central3-log1p-zscore-v1.1,,,,,,,390033274 -esm2_v1_1,18690,1280,float32,embeddings,data/embeddings/esm2_v1_1/esm2_v1_1_embeddings.npz,data/embeddings/esm2_v1_1/esm2_v1_1_genes.tsv,data/embeddings/esm2_v1_1/esm2_v1_1_metadata.json,,protein_sequence,ESM2-v1.1,,,,,,,88766138 -frogs_archs4_256_v1_1,16265,256,float32,embeddings,data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_embeddings.npz,data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_genes.tsv,data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_metadata.json,processed_qc_passed,,frogs_archs4_256_v1_1,,,,,,,15433953 -gene2vec_v1_1,17916,200,float32,embeddings,data/embeddings/gene2vec_v1_1/gene2vec_v1_1_embeddings.npz,data/embeddings/gene2vec_v1_1/gene2vec_v1_1_genes.tsv,data/embeddings/gene2vec_v1_1/gene2vec_v1_1_metadata.json,,coexpression_expression,Gene2Vec-v1.1,,,,,,,13292179 -genecompass_base_gene_token_v1_1,16578,768,float32,embeddings,data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_embeddings.npz,data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_genes.tsv,data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_metadata.json,processed_qc_passed,gene expression single-cell foundation model / knowledge-informed gene-token embedding,genecompass_base_gene_token_v1_1,official GeneCompass ScienceDB checkpoint linked from official GitHub README,"GeneCompass Base, 12-layer knowledge-informed cross-species foundation model",Daniel locally by extracting checkpoint embedding weights; no training or inference,,none; static learned token embedding weight extracted directly from checkpoint,,47272177 -geneformer_v2_316m_v1_1,20007,1152,float32,embeddings,data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_embeddings.npz,data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_genes.tsv,data/embeddings/geneformer_v2_316m_v1_1/geneformer_v2_316m_v1_1_metadata.json,,single_cell_expression_foundation_model,Geneformer-V2-316M-v1.1,Official ctheodoris/Geneformer Hugging Face repository,,Extracted locally from official Geneformer model input gene-token embedding layer,,,,85511948 -genept_ada_v1_1_conservative,19099,1536,float32,embeddings,data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_embeddings.npz,data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_genes.tsv,data/embeddings/genept_ada_v1_1_conservative/genept_ada_v1_1_conservative_metadata.json,,biomedical_literature_text,GenePT-ADA-v1.1-conservative,,,,,,,98626668 -genept_model3_gene_protein_v1_1_conservative,19103,3072,float32,embeddings,data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_embeddings.npz,data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_genes.tsv,data/embeddings/genept_model3_gene_protein_v1_1_conservative/genept_model3_gene_protein_v1_1_conservative_metadata.json,,biomedical_literature_text_plus_protein_sequence_text,GenePT-Model3-GeneProtein-v1.1-conservative,,,,,,,189551495 -gtex_tissue_median_log1p_zscore_v1_1,19353,68,float32,embeddings,data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_embeddings.npz,data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_genes.tsv,data/embeddings/gtex_tissue_median_log1p_zscore_v1_1/gtex_tissue_median_log1p_zscore_v1_1_metadata.json,processed_qc_passed,bulk RNA expression / normal tissue expression,GTEx-tissue-median-log1p-zscore-v1.1,,,,,,,4485371 -helix_mrna_full_transcript_v1_1,19976,256,float32,embeddings,data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_embeddings.npz,data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_genes.tsv,data/embeddings/helix_mrna_full_transcript_v1_1/helix_mrna_full_transcript_v1_1_metadata.json,processed_qc_passed,mRNA language model / mature transcript sequence,helix_mrna_full_transcript_v1_1,official HuggingFace model repository,Helix-mRNA,Daniel locally by model inference,helical-ai/helix-mRNA,"final token embeddings from official Helix wrapper, mean pooled over non-special sequence tokens; long transcripts chunked and token-count weighted",,19474797 -hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1,18787,256,float32,embeddings,data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_embeddings.npz,data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_genes.tsv,data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1/hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1_metadata.json,processed_qc_passed,heterogeneous biomedical knowledge graph / non-PPI gene context,Hetionet-nonPPI-biomedical-context-TFIDF-SVD256-v1.1,,,,,,,17704572 -hpa_normal_ihc_tissue_level_zscore_v1_1,13420,59,float32,embeddings,data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_embeddings.npz,data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_genes.tsv,data/embeddings/hpa_normal_ihc_tissue_level_zscore_v1_1/hpa_normal_ihc_tissue_level_zscore_v1_1_metadata.json,processed_qc_passed,protein expression / normal tissue immunohistochemistry,HPA-normal-IHC-tissue-level-zscore-v1.1,,,,,,,251036 -hpa_subcellular_location_multihot_zscore_v1_1,13581,50,float32,embeddings,data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_embeddings.npz,data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_genes.tsv,data/embeddings/hpa_subcellular_location_multihot_zscore_v1_1/hpa_subcellular_location_multihot_zscore_v1_1_metadata.json,processed_qc_passed,protein subcellular localization,HPA-subcellular-location-multihot-zscore-v1.1,,,,,,,38386 -hyenadna_medium_160k_cds_v1_1,20085,256,float32,embeddings,data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_embeddings.npz,data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_genes.tsv,data/embeddings/hyenadna_medium_160k_cds_v1_1/hyenadna_medium_160k_cds_v1_1_metadata.json,processed_qc_passed,DNA language model / canonical CDS sequence,hyenadna_medium_160k_cds_v1_1,official HuggingFace repository LongSafari/hyenadna-medium-160k-seqlen-hf,HyenaDNA medium 160k sequence-length checkpoint,Daniel locally by model inference,,final hidden-state mean pooling over non-special nucleotide tokens; long CDS split into chunks and token-count weighted,,18551316 -mahi_all_contexts_mean_v1_1,18918,3666,float32,embeddings,data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_embeddings.npz,data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_genes.tsv,data/embeddings/mahi_all_contexts_mean_v1_1/mahi_all_contexts_mean_v1_1_metadata.json,processed_qc_passed,,mahi_all_contexts_mean_v1_1,,,,,mean across all non-global tissue/cell contexts per Entrez gene,,245994508 -mahi_global_v1_1,18918,3666,float32,embeddings,data/embeddings/mahi_global_v1_1/mahi_global_v1_1_embeddings.npz,data/embeddings/mahi_global_v1_1/mahi_global_v1_1_genes.tsv,data/embeddings/mahi_global_v1_1/mahi_global_v1_1_metadata.json,processed_qc_passed,,mahi_global_v1_1,,,,,global context only,,167568851 -mashup_string_v1_1,16331,800,float32,embeddings,data/embeddings/mashup_string_v1_1/mashup_string_v1_1_embeddings.npz,data/embeddings/mashup_string_v1_1/mashup_string_v1_1_genes.tsv,data/embeddings/mashup_string_v1_1/mashup_string_v1_1_metadata.json,,network_ppi,Mashup-STRING-v1.1,,,,,,,47906054 -msa_transformer_depth1_uniprot_human_v1_1,19472,768,float32,embeddings,data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_embeddings.npz,data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_genes.tsv,data/embeddings/msa_transformer_depth1_uniprot_human_v1_1/msa_transformer_depth1_uniprot_human_v1_1_metadata.json,,protein_sequence_msa_transformer_depth1,MSATransformer-depth1-UniProt-human-v1.1,,MSA Transformer,Daniel locally,esm_msa1b_t12_100M_UR50S,mean residue representation from layer 12; long proteins split into non-overlapping windows and length-weighted mean pooled,,55476461 -newt_archs4_256_v1_1,16265,256,float32,embeddings,data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_embeddings.npz,data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_genes.tsv,data/embeddings/newt_archs4_256_v1_1/newt_archs4_256_v1_1_metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-ARCHS4-256-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,15434849 -newt_cellnet_v1_1,17499,128,float32,embeddings,data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_embeddings.npz,data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_genes.tsv,data/embeddings/newt_cellnet_v1_1/newt_cellnet_v1_1_metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-CellNet-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,8320324 -newt_go_256_v1_1,13788,256,float32,embeddings,data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_embeddings.npz,data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_genes.tsv,data/embeddings/newt_go_256_v1_1/newt_go_256_v1_1_metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-GO-256-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,13109399 -newt_go_graph_v1_1,13788,256,float32,embeddings,data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_embeddings.npz,data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_genes.tsv,data/embeddings/newt_go_graph_v1_1/newt_go_graph_v1_1_metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-GO-graph-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,13142698 -newt_msigdb_bundle_v1_1,15423,256,float32,embeddings,data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_embeddings.npz,data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_genes.tsv,data/embeddings/newt_msigdb_bundle_v1_1/newt_msigdb_bundle_v1_1_metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-MSigDB-bundle-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,14773343 -node2vec_consensus_ppi_v1_1,18356,128,float32,embeddings,data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_embeddings.npz,data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_genes.tsv,data/embeddings/node2vec_consensus_ppi_v1_1/node2vec_consensus_ppi_v1_1_metadata.json,,PPI network embedding,Node2Vec-consensus-PPI-v1.1,ylaboratory/gene-embedding-benchmarks GitHub repository: src/other/preprocess_embedding/consensus.dat,node2vec: scalable feature learning for networks,Daniel locally from consensus.dat using reproducible random walks and Word2Vec skip-gram,,,,8710233 -nt_v2_500m_multispecies_cds_v1_1,20085,1024,float32,embeddings,data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_embeddings.npz,data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_genes.tsv,data/embeddings/nt_v2_500m_multispecies_cds_v1_1/nt_v2_500m_multispecies_cds_v1_1_metadata.json,processed_qc_passed,DNA language model / CDS sequence,nt_v2_500m_multispecies_cds_v1_1,official HuggingFace model repository,Nucleotide Transformer v2,Daniel locally by model inference,InstaDeepAI/nucleotide-transformer-v2-500m-multi-species,"final hidden layer mean pooling over non-special, non-padding tokens; chunk-weighted by token count for long CDS",,76371728 -opa2vec_goa_human_v1_1,18660,200,float32,embeddings,data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_embeddings.npz,data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_genes.tsv,data/embeddings/opa2vec_goa_human_v1_1/opa2vec_goa_human_v1_1_metadata.json,,Gene Ontology / ontology-annotation semantic embedding,OPA2Vec-GOA-human-v1.1,Gene Ontology go-basic.obo and GOA human GAF,OPA2Vec,Daniel locally using an OPA2Vec-derived Python 3 workflow,,,,13852912 -orthrus_base_4track_full_transcript_v1_1,19976,256,float32,embeddings,data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_embeddings.npz,data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_genes.tsv,data/embeddings/orthrus_base_4track_full_transcript_v1_1/orthrus_base_4track_full_transcript_v1_1_metadata.json,processed_qc_passed,,orthrus_base_4track_full_transcript_v1_1,,,,,"Orthrus representation() sequence-level embedding. For transcripts longer than max_full_len, length-weighted mean of chunk-level Orthrus embeddings.",,18950246 -primekg_nonppi_biomedical_context_tfidf_svd256_v1_1,19216,256,float32,embeddings,data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_embeddings.npz,data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_genes.tsv,data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1/primekg_nonppi_biomedical_context_tfidf_svd256_v1_1_metadata.json,processed_qc_passed,heterogeneous biomedical knowledge graph / non-PPI gene context,PrimeKG-nonPPI-biomedical-context-TFIDF-SVD256-v1.1,,,,,,,18076131 -probe_aac_v1_1,18873,20,float32,embeddings,data/embeddings/probe_aac_v1_1/probe_aac_v1_1_embeddings.npz,data/embeddings/probe_aac_v1_1/probe_aac_v1_1_genes.tsv,data/embeddings/probe_aac_v1_1/probe_aac_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_aac_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,AAC,PROBE authors / original PROBE human representation archive,,,,982437 -probe_albert_v1_1,18453,4096,float32,embeddings,data/embeddings/probe_albert_v1_1/probe_albert_v1_1_embeddings.npz,data/embeddings/probe_albert_v1_1/probe_albert_v1_1_genes.tsv,data/embeddings/probe_albert_v1_1/probe_albert_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_albert_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,ALBERT,PROBE authors / original PROBE human representation archive,,,,280084766 -probe_apaac_v1_1,18871,80,float32,embeddings,data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_embeddings.npz,data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_genes.tsv,data/embeddings/probe_apaac_v1_1/probe_apaac_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_apaac_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,APAAC,PROBE authors / original PROBE human representation archive,,,,5430243 -probe_bert_bfd_v1_1,18453,1024,float32,embeddings,data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_embeddings.npz,data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_genes.tsv,data/embeddings/probe_bert_bfd_v1_1/probe_bert_bfd_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_bert_bfd_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,BERT-BFD,PROBE authors / original PROBE human representation archive,,,,70423891 -probe_bert_pfam_v1_1,18461,768,float32,embeddings,data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_embeddings.npz,data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_genes.tsv,data/embeddings/probe_bert_pfam_v1_1/probe_bert_pfam_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_bert_pfam_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,BERT-PFAM,PROBE authors / original PROBE human representation archive,,,,52703247 -probe_blast_v1_1,18873,20421,float32,embeddings,data/embeddings/probe_blast_v1_1/probe_blast_v1_1_embeddings.npz,data/embeddings/probe_blast_v1_1/probe_blast_v1_1_genes.tsv,data/embeddings/probe_blast_v1_1/probe_blast_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_blast_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,BLAST,PROBE authors / original PROBE human representation archive,,,,5517015 -probe_cpc_prot_v1_1,18453,512,float32,embeddings,data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_embeddings.npz,data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_genes.tsv,data/embeddings/probe_cpc_prot_v1_1/probe_cpc_prot_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_cpc_prot_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,CPC-PROT,PROBE authors / original PROBE human representation archive,,,,29870283 -probe_esmb1_v1_1,18453,1280,float32,embeddings,data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_embeddings.npz,data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_genes.tsv,data/embeddings/probe_esmb1_v1_1/probe_esmb1_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_esmb1_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,ESMB1,PROBE authors / original PROBE human representation archive,,,,87778870 -probe_gene2vec_uniprot_v1_1,17564,200,float32,embeddings,data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_embeddings.npz,data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_genes.tsv,data/embeddings/probe_gene2vec_uniprot_v1_1/probe_gene2vec_uniprot_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_gene2vec_uniprot_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,GENE2VEC,PROBE authors / original PROBE human representation archive,,,,13029054 -probe_hmmer_v1_1,18873,20421,float32,embeddings,data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_embeddings.npz,data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_genes.tsv,data/embeddings/probe_hmmer_v1_1/probe_hmmer_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_hmmer_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,HMMER,PROBE authors / original PROBE human representation archive,,,,18898149 -probe_ksep_v1_1,18801,400,float32,embeddings,data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_embeddings.npz,data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_genes.tsv,data/embeddings/probe_ksep_v1_1/probe_ksep_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_ksep_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,KSEP,PROBE authors / original PROBE human representation archive,,,,23775898 -probe_learned_vec_v1_1,18873,64,float32,embeddings,data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_embeddings.npz,data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_genes.tsv,data/embeddings/probe_learned_vec_v1_1/probe_learned_vec_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_learned_vec_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,LEARNED-VEC,PROBE authors / original PROBE human representation archive,,,,4489189 -probe_mut2vec_v1_1,17445,300,float32,embeddings,data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_embeddings.npz,data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_genes.tsv,data/embeddings/probe_mut2vec_v1_1/probe_mut2vec_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_mut2vec_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,MUT2VEC,PROBE authors / original PROBE human representation archive,,,,10844260 -probe_pfam_v1_1,18873,6227,float32,embeddings,data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_embeddings.npz,data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_genes.tsv,data/embeddings/probe_pfam_v1_1/probe_pfam_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_pfam_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,PFAM,PROBE authors / original PROBE human representation archive,,,,548968 -probe_protvec_v1_1,18873,100,float32,embeddings,data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_embeddings.npz,data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_genes.tsv,data/embeddings/probe_protvec_v1_1/probe_protvec_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_protvec_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,PROTVEC,PROBE authors / original PROBE human representation archive,,,,7066137 -probe_seqvec_v1_1,18873,1024,float32,embeddings,data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_embeddings.npz,data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_genes.tsv,data/embeddings/probe_seqvec_v1_1/probe_seqvec_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_seqvec_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,SEQVEC,PROBE authors / original PROBE human representation archive,,,,71841293 -probe_t5_v1_1,18453,1024,float32,embeddings,data/embeddings/probe_t5_v1_1/probe_t5_v1_1_embeddings.npz,data/embeddings/probe_t5_v1_1/probe_t5_v1_1_genes.tsv,data/embeddings/probe_t5_v1_1/probe_t5_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_t5_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,T5,PROBE authors / original PROBE human representation archive,,,,70153727 -probe_tcga_embedding_v1_1,18221,50,float32,embeddings,data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_embeddings.npz,data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_genes.tsv,data/embeddings/probe_tcga_embedding_v1_1/probe_tcga_embedding_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_tcga_embedding_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,TCGA-EMBEDDING,PROBE authors / original PROBE human representation archive,,,,3378621 -probe_unirep_v1_1,18873,5700,float32,embeddings,data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_embeddings.npz,data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_genes.tsv,data/embeddings/probe_unirep_v1_1/probe_unirep_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_unirep_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,UNIREP,PROBE authors / original PROBE human representation archive,,,,410932244 -probe_xlnet_v1_1,18453,1024,float32,embeddings,data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_embeddings.npz,data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_genes.tsv,data/embeddings/probe_xlnet_v1_1/probe_xlnet_v1_1_metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_xlnet_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,XLNET,PROBE authors / original PROBE human representation archive,,,,70519787 -pubmedbert_gene_protein_summary_v1_1,19168,768,float32,embeddings,data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_embeddings.npz,data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_genes.tsv,data/embeddings/pubmedbert_gene_protein_summary_v1_1/pubmedbert_gene_protein_summary_v1_1_metadata.json,,literature_text_gene_summary,PubMedBERT-GeneProteinSummary-v1.1,GenePT Zenodo v2 summaries,,Daniel locally,microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext,"final hidden layer mean pooling over valid non-special, non-padding tokens",,54786752 -rnafm_full_transcript_v1_1,19976,640,float32,embeddings,data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_embeddings.npz,data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_genes.tsv,data/embeddings/rnafm_full_transcript_v1_1/rnafm_full_transcript_v1_1_metadata.json,processed_qc_passed,,rnafm_full_transcript_v1_1,,,,/local/home/dtabares/data/raw_embeddings/rnafm_full_transcript_v1_1/hf_model/RNA-FM_pretrained.pth,Final layer 12 mean token pooling over real RNA tokens only; BOS/EOS excluded; long transcripts split into chunks and chunk means length-weighted,,47521441 -scfoundation_gene_pos_v1_1,19062,768,float32,embeddings,data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_embeddings.npz,data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_genes.tsv,data/embeddings/scfoundation_gene_pos_v1_1/scfoundation_gene_pos_v1_1_metadata.json,,single_cell_expression_foundation_model / learned gene identity-position embedding,scFoundation-gene-pos-v1.1,Official scFoundation GitHub repository and official SharePoint checkpoint linked from the repository,scFoundation / xTrimoGene,Extracted locally from official scFoundation gene sub-checkpoint model.pos_emb.weight,,,,54234742 -scgpt_pancancer_v1_1,19071,512,float32,embeddings,data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_embeddings.npz,data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_genes.tsv,data/embeddings/scgpt_pancancer_v1_1/scgpt_pancancer_v1_1_metadata.json,processed_qc_passed,,scgpt_pancancer_v1_1,,,,,,,36187165 -scgpt_whole_human_v1_1,19280,512,float32,embeddings,data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_embeddings.npz,data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_genes.tsv,data/embeddings/scgpt_whole_human_v1_1/scgpt_whole_human_v1_1_metadata.json,,single_cell_expression_foundation_model,scGPT-whole-human-v1.1,Official scGPT whole-human Google Drive checkpoint linked from the scGPT project,,Extracted locally from official scGPT whole-human model input gene-token embedding layer,,,,36573474 -scprint_medium_v1_5_gene_v1_1,19995,256,float32,embeddings,data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_embeddings.npz,data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_genes.tsv,data/embeddings/scprint_medium_v1_5_gene_v1_1/scprint_medium_v1_5_gene_v1_1_metadata.json,,single_cell_expression_foundation_model / learned gene-token embedding,scPRINT-medium-v1.5-gene-v1.1,Official scPRINT HuggingFace repository jkobject/scPRINT,scPRINT,Extracted locally from official scPRINT medium-v1.5 checkpoint,,,,18989317 -spliceai_canonical_splice_site_profile_v1_1,20097,606,float32,embeddings,data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_embeddings.npz,data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_genes.tsv,data/embeddings/spliceai_canonical_splice_site_profile_v1_1/spliceai_canonical_splice_site_profile_v1_1_metadata.json,processed_qc_passed,splicing sequence model / canonical transcript splice-site profiles,SpliceAI-canonical-splice-site-profile-zscore-v1.1,,,,,,,41717181 -uce_33l_v1_1,19082,1280,float32,embeddings,data/embeddings/uce_33l_v1_1/uce_33l_v1_1_embeddings.npz,data/embeddings/uce_33l_v1_1/uce_33l_v1_1_genes.tsv,data/embeddings/uce_33l_v1_1/uce_33l_v1_1_metadata.json,,single_cell_expression_foundation_model / protein-informed gene-token embedding,UCE-33L-v1.1,Official UCE GitHub repository plus official Figshare model files,Universal Cell Embeddings (UCE),Extracted locally from official UCE all_tokens and 33-layer checkpoint gene encoder,,,,89725640 +former_processing_embedding_id,embedding_id,n_genes,dimension,dtype,npz_array_key,embeddings_npz,genes_tsv,metadata_json,status,modality,embedding_name,download_source,original_model_or_method,embedding_generated_by,model_checkpoint,pooling_strategy,license,file_size_bytes +alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1,alphagenome_tss1mb_regulatory_tracksummary_pca256,20084,256,float32,embeddings,data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/embeddings.npz,data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/genes.tsv,data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/metadata.json,processed_qc_passed,,AlphaGenome-TSS1Mb-regulatoryTrackSummary-PCA256-v1.1,,,,,,,19270344 +basenji2_tss_131kb_human_central3_v1_1,basenji2_tss_131kb_human_central3,20097,5313,float32,embeddings,data/embeddings/basenji2_tss_131kb_human_central3/embeddings.npz,data/embeddings/basenji2_tss_131kb_human_central3/genes.tsv,data/embeddings/basenji2_tss_131kb_human_central3/metadata.json,processed_qc_passed,regulatory DNA / Basenji2 TSS sequence model,Basenji2-TSS-131kb-human-central3-log1p-zscore-v1.1,,,,,,,389641898 +biobert_gene_protein_summary_v1_1,biobert_gene_protein_summary,19168,768,float32,embeddings,data/embeddings/biobert_gene_protein_summary/embeddings.npz,data/embeddings/biobert_gene_protein_summary/genes.tsv,data/embeddings/biobert_gene_protein_summary/metadata.json,,literature_text_gene_summary,BioBERT-GeneProteinSummary-v1.1,GenePT Zenodo v2 summaries,,Daniel locally,dmis-lab/biobert-base-cased-v1.2,"final hidden layer mean pooling over valid non-special, non-padding tokens",,54606212 +bioconceptvec_cbow_v1_1,bioconceptvec_cbow,17040,100,float32,embeddings,data/embeddings/bioconceptvec_cbow/embeddings.npz,data/embeddings/bioconceptvec_cbow/genes.tsv,data/embeddings/bioconceptvec_cbow/metadata.json,,biomedical_literature_text,BioConceptVec-CBOW-v1.1,,,,,,,6339061 +bioconceptvec_fasttext_v1_1,bioconceptvec_fasttext,17040,100,float32,embeddings,data/embeddings/bioconceptvec_fasttext/embeddings.npz,data/embeddings/bioconceptvec_fasttext/genes.tsv,data/embeddings/bioconceptvec_fasttext/metadata.json,,biomedical_literature_text,BioConceptVec-fastText-v1.1,,,,,,,6254929 +bioconceptvec_glove_v1_1,bioconceptvec_glove,17040,100,float32,embeddings,data/embeddings/bioconceptvec_glove/embeddings.npz,data/embeddings/bioconceptvec_glove/genes.tsv,data/embeddings/bioconceptvec_glove/metadata.json,,biomedical_literature_text,BioConceptVec-GloVe-v1.1,,,,,,,6350146 +bioconceptvec_skipgram_v1_1,bioconceptvec_skipgram,17040,100,float32,embeddings,data/embeddings/bioconceptvec_skipgram/embeddings.npz,data/embeddings/bioconceptvec_skipgram/genes.tsv,data/embeddings/bioconceptvec_skipgram/metadata.json,,biomedical_literature_text,BioConceptVec-SkipGram-v1.1,,,,,,,6335039 +cmap_gse106127_cgs_mean_landmark978_zscore_v1_1,cmap_gse106127_cgs_mean_landmark978_zscore,4070,978,float32,embeddings,data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/embeddings.npz,data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/genes.tsv,data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/metadata.json,processed_qc_passed,perturbational transcriptomics / L1000 consensus genetic perturbation signatures,CMap-GSE106127-CGS-mean-landmark978-zscore-v1.1,,,,,,,14768959 +deepnf_string_v12_ppmi_svd_fusion_v1_1,deepnf_string_v12_ppmi_svd_fusion,19071,256,float32,embeddings,data/embeddings/deepnf_string_v12_ppmi_svd_fusion/embeddings.npz,data/embeddings/deepnf_string_v12_ppmi_svd_fusion/genes.tsv,data/embeddings/deepnf_string_v12_ppmi_svd_fusion/metadata.json,,STRING network multimodal sparse PPMI/SVD fusion,deepNF-STRING-v12-PPMI-SVD-fusion-v1.1,"STRING v12 human protein.links.detailed, protein.info, and protein.aliases files",deepNF: deep network fusion for protein function prediction,Daniel locally from STRING v12 channel networks,,,,18225284 +deepsea_tss_1kb_epigenomic_probs_v1_1,deepsea_tss_1kb_epigenomic_probs,20097,919,float32,embeddings,data/embeddings/deepsea_tss_1kb_epigenomic_probs/embeddings.npz,data/embeddings/deepsea_tss_1kb_epigenomic_probs/genes.tsv,data/embeddings/deepsea_tss_1kb_epigenomic_probs/metadata.json,processed_qc_passed,regulatory DNA / promoter-TSS sequence model,DeepSEA-TSS-1kb-epigenomic-probabilities-zscore-v1.1,,,,,,,66380222 +depmap_crispr_gene_effect_pca256_v1_1,depmap_crispr_gene_effect_pca256,18562,256,float32,embeddings,data/embeddings/depmap_crispr_gene_effect_pca256/embeddings.npz,data/embeddings/depmap_crispr_gene_effect_pca256/genes.tsv,data/embeddings/depmap_crispr_gene_effect_pca256/metadata.json,processed_qc_passed,functional genomics / CRISPR dependency,DepMap-CRISPRGeneEffect-PCA256-v1.1,,,,,,,17729467 +dnabert2_117m_cds_v1_1,dnabert2_117m_cds,20085,768,float32,embeddings,data/embeddings/dnabert2_117m_cds/embeddings.npz,data/embeddings/dnabert2_117m_cds/genes.tsv,data/embeddings/dnabert2_117m_cds/metadata.json,processed_qc_passed,,dnabert2_117m_cds_v1_1,,,,,"Mean pooling over final-layer non-special token embeddings. Long CDS sequences split into nucleotide chunks, then length-weighted mean of chunk-level embeddings.",,57063502 +enformer_tss_393kb_human_central3_v1_1,enformer_tss_393kb_human_central3,20097,5313,float32,embeddings,data/embeddings/enformer_tss_393kb_human_central3/embeddings.npz,data/embeddings/enformer_tss_393kb_human_central3/genes.tsv,data/embeddings/enformer_tss_393kb_human_central3/metadata.json,processed_qc_passed,regulatory DNA / long-range TSS sequence model,Enformer-TSS-393kb-human-central3-log1p-zscore-v1.1,,,,,,,390033274 +esm2_v1_1,esm2,18690,1280,float32,embeddings,data/embeddings/esm2/embeddings.npz,data/embeddings/esm2/genes.tsv,data/embeddings/esm2/metadata.json,,protein_sequence,ESM2-v1.1,,,,,,,88766138 +frogs_archs4_256_v1_1,frogs_archs4_256,16265,256,float32,embeddings,data/embeddings/frogs_archs4_256/embeddings.npz,data/embeddings/frogs_archs4_256/genes.tsv,data/embeddings/frogs_archs4_256/metadata.json,processed_qc_passed,,frogs_archs4_256_v1_1,,,,,,,15433953 +gene2vec_v1_1,gene2vec,17916,200,float32,embeddings,data/embeddings/gene2vec/embeddings.npz,data/embeddings/gene2vec/genes.tsv,data/embeddings/gene2vec/metadata.json,,coexpression_expression,Gene2Vec-v1.1,,,,,,,13292179 +genecompass_base_gene_token_v1_1,genecompass_base_gene_token,16578,768,float32,embeddings,data/embeddings/genecompass_base_gene_token/embeddings.npz,data/embeddings/genecompass_base_gene_token/genes.tsv,data/embeddings/genecompass_base_gene_token/metadata.json,processed_qc_passed,gene expression single-cell foundation model / knowledge-informed gene-token embedding,genecompass_base_gene_token_v1_1,official GeneCompass ScienceDB checkpoint linked from official GitHub README,"GeneCompass Base, 12-layer knowledge-informed cross-species foundation model",Daniel locally by extracting checkpoint embedding weights; no training or inference,,none; static learned token embedding weight extracted directly from checkpoint,,47272177 +geneformer_v2_316m_v1_1,geneformer_v2_316m,20007,1152,float32,embeddings,data/embeddings/geneformer_v2_316m/embeddings.npz,data/embeddings/geneformer_v2_316m/genes.tsv,data/embeddings/geneformer_v2_316m/metadata.json,,single_cell_expression_foundation_model,Geneformer-V2-316M-v1.1,Official ctheodoris/Geneformer Hugging Face repository,,Extracted locally from official Geneformer model input gene-token embedding layer,,,,85511948 +genept_ada_v1_1_conservative,genept_ada_conservative,19099,1536,float32,embeddings,data/embeddings/genept_ada_conservative/embeddings.npz,data/embeddings/genept_ada_conservative/genes.tsv,data/embeddings/genept_ada_conservative/metadata.json,,biomedical_literature_text,GenePT-ADA-v1.1-conservative,,,,,,,98626668 +genept_model3_gene_protein_v1_1_conservative,genept_model3_gene_protein_conservative,19103,3072,float32,embeddings,data/embeddings/genept_model3_gene_protein_conservative/embeddings.npz,data/embeddings/genept_model3_gene_protein_conservative/genes.tsv,data/embeddings/genept_model3_gene_protein_conservative/metadata.json,,biomedical_literature_text_plus_protein_sequence_text,GenePT-Model3-GeneProtein-v1.1-conservative,,,,,,,189551495 +gtex_tissue_median_log1p_zscore_v1_1,gtex_tissue_median_log1p_zscore,19353,68,float32,embeddings,data/embeddings/gtex_tissue_median_log1p_zscore/embeddings.npz,data/embeddings/gtex_tissue_median_log1p_zscore/genes.tsv,data/embeddings/gtex_tissue_median_log1p_zscore/metadata.json,processed_qc_passed,bulk RNA expression / normal tissue expression,GTEx-tissue-median-log1p-zscore-v1.1,,,,,,,4485371 +helix_mrna_full_transcript_v1_1,helix_mrna_full_transcript,19976,256,float32,embeddings,data/embeddings/helix_mrna_full_transcript/embeddings.npz,data/embeddings/helix_mrna_full_transcript/genes.tsv,data/embeddings/helix_mrna_full_transcript/metadata.json,processed_qc_passed,mRNA language model / mature transcript sequence,helix_mrna_full_transcript_v1_1,official HuggingFace model repository,Helix-mRNA,Daniel locally by model inference,helical-ai/helix-mRNA,"final token embeddings from official Helix wrapper, mean pooled over non-special sequence tokens; long transcripts chunked and token-count weighted",,19474797 +hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1,hetionet_nonppi_biomedical_context_tfidf_svd256,18787,256,float32,embeddings,data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/embeddings.npz,data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/genes.tsv,data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/metadata.json,processed_qc_passed,heterogeneous biomedical knowledge graph / non-PPI gene context,Hetionet-nonPPI-biomedical-context-TFIDF-SVD256-v1.1,,,,,,,17704572 +hpa_normal_ihc_tissue_level_zscore_v1_1,hpa_normal_ihc_tissue_level_zscore,13420,59,float32,embeddings,data/embeddings/hpa_normal_ihc_tissue_level_zscore/embeddings.npz,data/embeddings/hpa_normal_ihc_tissue_level_zscore/genes.tsv,data/embeddings/hpa_normal_ihc_tissue_level_zscore/metadata.json,processed_qc_passed,protein expression / normal tissue immunohistochemistry,HPA-normal-IHC-tissue-level-zscore-v1.1,,,,,,,251036 +hpa_subcellular_location_multihot_zscore_v1_1,hpa_subcellular_location_multihot_zscore,13581,50,float32,embeddings,data/embeddings/hpa_subcellular_location_multihot_zscore/embeddings.npz,data/embeddings/hpa_subcellular_location_multihot_zscore/genes.tsv,data/embeddings/hpa_subcellular_location_multihot_zscore/metadata.json,processed_qc_passed,protein subcellular localization,HPA-subcellular-location-multihot-zscore-v1.1,,,,,,,38386 +hyenadna_medium_160k_cds_v1_1,hyenadna_medium_160k_cds,20085,256,float32,embeddings,data/embeddings/hyenadna_medium_160k_cds/embeddings.npz,data/embeddings/hyenadna_medium_160k_cds/genes.tsv,data/embeddings/hyenadna_medium_160k_cds/metadata.json,processed_qc_passed,DNA language model / canonical CDS sequence,hyenadna_medium_160k_cds_v1_1,official HuggingFace repository LongSafari/hyenadna-medium-160k-seqlen-hf,HyenaDNA medium 160k sequence-length checkpoint,Daniel locally by model inference,,final hidden-state mean pooling over non-special nucleotide tokens; long CDS split into chunks and token-count weighted,,18551316 +mahi_all_contexts_mean_v1_1,mahi_all_contexts_mean,18918,3666,float32,embeddings,data/embeddings/mahi_all_contexts_mean/embeddings.npz,data/embeddings/mahi_all_contexts_mean/genes.tsv,data/embeddings/mahi_all_contexts_mean/metadata.json,processed_qc_passed,,mahi_all_contexts_mean_v1_1,,,,,mean across all non-global tissue/cell contexts per Entrez gene,,245994508 +mahi_global_v1_1,mahi_global,18918,3666,float32,embeddings,data/embeddings/mahi_global/embeddings.npz,data/embeddings/mahi_global/genes.tsv,data/embeddings/mahi_global/metadata.json,processed_qc_passed,,mahi_global_v1_1,,,,,global context only,,167568851 +mashup_string_v1_1,mashup_string,16331,800,float32,embeddings,data/embeddings/mashup_string/embeddings.npz,data/embeddings/mashup_string/genes.tsv,data/embeddings/mashup_string/metadata.json,,network_ppi,Mashup-STRING-v1.1,,,,,,,47906054 +msa_transformer_depth1_uniprot_human_v1_1,msa_transformer_depth1_uniprot_human,19472,768,float32,embeddings,data/embeddings/msa_transformer_depth1_uniprot_human/embeddings.npz,data/embeddings/msa_transformer_depth1_uniprot_human/genes.tsv,data/embeddings/msa_transformer_depth1_uniprot_human/metadata.json,,protein_sequence_msa_transformer_depth1,MSATransformer-depth1-UniProt-human-v1.1,,MSA Transformer,Daniel locally,esm_msa1b_t12_100M_UR50S,mean residue representation from layer 12; long proteins split into non-overlapping windows and length-weighted mean pooled,,55476461 +newt_archs4_256_v1_1,newt_archs4_256,16265,256,float32,embeddings,data/embeddings/newt_archs4_256/embeddings.npz,data/embeddings/newt_archs4_256/genes.tsv,data/embeddings/newt_archs4_256/metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-ARCHS4-256-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,15434849 +newt_cellnet_v1_1,newt_cellnet,17499,128,float32,embeddings,data/embeddings/newt_cellnet/embeddings.npz,data/embeddings/newt_cellnet/genes.tsv,data/embeddings/newt_cellnet/metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-CellNet-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,8320324 +newt_go_256_v1_1,newt_go_256,13788,256,float32,embeddings,data/embeddings/newt_go_256/embeddings.npz,data/embeddings/newt_go_256/genes.tsv,data/embeddings/newt_go_256/metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-GO-256-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,13109399 +newt_go_graph_v1_1,newt_go_graph,13788,256,float32,embeddings,data/embeddings/newt_go_graph/embeddings.npz,data/embeddings/newt_go_graph/genes.tsv,data/embeddings/newt_go_graph/metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-GO-graph-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,13142698 +newt_msigdb_bundle_v1_1,newt_msigdb_bundle,15423,256,float32,embeddings,data/embeddings/newt_msigdb_bundle/embeddings.npz,data/embeddings/newt_msigdb_bundle/genes.tsv,data/embeddings/newt_msigdb_bundle/metadata.json,,multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT,NEWT-MSigDB-bundle-v1.1,"Official NEWT Zenodo dataset release, DOI 10.5281/zenodo.19340318",NEWT component embedding,NEWT authors; processed locally into the project standard format,,,,14773343 +node2vec_consensus_ppi_v1_1,node2vec_consensus_ppi,18356,128,float32,embeddings,data/embeddings/node2vec_consensus_ppi/embeddings.npz,data/embeddings/node2vec_consensus_ppi/genes.tsv,data/embeddings/node2vec_consensus_ppi/metadata.json,,PPI network embedding,Node2Vec-consensus-PPI-v1.1,ylaboratory/gene-embedding-benchmarks GitHub repository: src/other/preprocess_embedding/consensus.dat,node2vec: scalable feature learning for networks,Daniel locally from consensus.dat using reproducible random walks and Word2Vec skip-gram,,,,8710233 +nt_v2_500m_multispecies_cds_v1_1,nt_v2_500m_multispecies_cds,20085,1024,float32,embeddings,data/embeddings/nt_v2_500m_multispecies_cds/embeddings.npz,data/embeddings/nt_v2_500m_multispecies_cds/genes.tsv,data/embeddings/nt_v2_500m_multispecies_cds/metadata.json,processed_qc_passed,DNA language model / CDS sequence,nt_v2_500m_multispecies_cds_v1_1,official HuggingFace model repository,Nucleotide Transformer v2,Daniel locally by model inference,InstaDeepAI/nucleotide-transformer-v2-500m-multi-species,"final hidden layer mean pooling over non-special, non-padding tokens; chunk-weighted by token count for long CDS",,76371728 +opa2vec_goa_human_v1_1,opa2vec_goa_human,18660,200,float32,embeddings,data/embeddings/opa2vec_goa_human/embeddings.npz,data/embeddings/opa2vec_goa_human/genes.tsv,data/embeddings/opa2vec_goa_human/metadata.json,,Gene Ontology / ontology-annotation semantic embedding,OPA2Vec-GOA-human-v1.1,Gene Ontology go-basic.obo and GOA human GAF,OPA2Vec,Daniel locally using an OPA2Vec-derived Python 3 workflow,,,,13852912 +orthrus_base_4track_full_transcript_v1_1,orthrus_base_4track_full_transcript,19976,256,float32,embeddings,data/embeddings/orthrus_base_4track_full_transcript/embeddings.npz,data/embeddings/orthrus_base_4track_full_transcript/genes.tsv,data/embeddings/orthrus_base_4track_full_transcript/metadata.json,processed_qc_passed,,orthrus_base_4track_full_transcript_v1_1,,,,,"Orthrus representation() sequence-level embedding. For transcripts longer than max_full_len, length-weighted mean of chunk-level Orthrus embeddings.",,18950246 +primekg_nonppi_biomedical_context_tfidf_svd256_v1_1,primekg_nonppi_biomedical_context_tfidf_svd256,19216,256,float32,embeddings,data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/embeddings.npz,data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/genes.tsv,data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/metadata.json,processed_qc_passed,heterogeneous biomedical knowledge graph / non-PPI gene context,PrimeKG-nonPPI-biomedical-context-TFIDF-SVD256-v1.1,,,,,,,18076131 +probe_aac_v1_1,probe_aac,18873,20,float32,embeddings,data/embeddings/probe_aac/embeddings.npz,data/embeddings/probe_aac/genes.tsv,data/embeddings/probe_aac/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_aac_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,AAC,PROBE authors / original PROBE human representation archive,,,,982437 +probe_albert_v1_1,probe_albert,18453,4096,float32,embeddings,data/embeddings/probe_albert/embeddings.npz,data/embeddings/probe_albert/genes.tsv,data/embeddings/probe_albert/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_albert_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,ALBERT,PROBE authors / original PROBE human representation archive,,,,280084766 +probe_apaac_v1_1,probe_apaac,18871,80,float32,embeddings,data/embeddings/probe_apaac/embeddings.npz,data/embeddings/probe_apaac/genes.tsv,data/embeddings/probe_apaac/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_apaac_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,APAAC,PROBE authors / original PROBE human representation archive,,,,5430243 +probe_bert_bfd_v1_1,probe_bert_bfd,18453,1024,float32,embeddings,data/embeddings/probe_bert_bfd/embeddings.npz,data/embeddings/probe_bert_bfd/genes.tsv,data/embeddings/probe_bert_bfd/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_bert_bfd_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,BERT-BFD,PROBE authors / original PROBE human representation archive,,,,70423891 +probe_bert_pfam_v1_1,probe_bert_pfam,18461,768,float32,embeddings,data/embeddings/probe_bert_pfam/embeddings.npz,data/embeddings/probe_bert_pfam/genes.tsv,data/embeddings/probe_bert_pfam/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_bert_pfam_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,BERT-PFAM,PROBE authors / original PROBE human representation archive,,,,52703247 +probe_blast_v1_1,probe_blast,18873,20421,float32,embeddings,data/embeddings/probe_blast/embeddings.npz,data/embeddings/probe_blast/genes.tsv,data/embeddings/probe_blast/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_blast_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,BLAST,PROBE authors / original PROBE human representation archive,,,,5517015 +probe_cpc_prot_v1_1,probe_cpc_prot,18453,512,float32,embeddings,data/embeddings/probe_cpc_prot/embeddings.npz,data/embeddings/probe_cpc_prot/genes.tsv,data/embeddings/probe_cpc_prot/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_cpc_prot_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,CPC-PROT,PROBE authors / original PROBE human representation archive,,,,29870283 +probe_esmb1_v1_1,probe_esmb1,18453,1280,float32,embeddings,data/embeddings/probe_esmb1/embeddings.npz,data/embeddings/probe_esmb1/genes.tsv,data/embeddings/probe_esmb1/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_esmb1_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,ESMB1,PROBE authors / original PROBE human representation archive,,,,87778870 +probe_gene2vec_uniprot_v1_1,probe_gene2vec_uniprot,17564,200,float32,embeddings,data/embeddings/probe_gene2vec_uniprot/embeddings.npz,data/embeddings/probe_gene2vec_uniprot/genes.tsv,data/embeddings/probe_gene2vec_uniprot/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_gene2vec_uniprot_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,GENE2VEC,PROBE authors / original PROBE human representation archive,,,,13029054 +probe_hmmer_v1_1,probe_hmmer,18873,20421,float32,embeddings,data/embeddings/probe_hmmer/embeddings.npz,data/embeddings/probe_hmmer/genes.tsv,data/embeddings/probe_hmmer/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_hmmer_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,HMMER,PROBE authors / original PROBE human representation archive,,,,18898149 +probe_ksep_v1_1,probe_ksep,18801,400,float32,embeddings,data/embeddings/probe_ksep/embeddings.npz,data/embeddings/probe_ksep/genes.tsv,data/embeddings/probe_ksep/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_ksep_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,KSEP,PROBE authors / original PROBE human representation archive,,,,23775898 +probe_learned_vec_v1_1,probe_learned_vec,18873,64,float32,embeddings,data/embeddings/probe_learned_vec/embeddings.npz,data/embeddings/probe_learned_vec/genes.tsv,data/embeddings/probe_learned_vec/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_learned_vec_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,LEARNED-VEC,PROBE authors / original PROBE human representation archive,,,,4489189 +probe_mut2vec_v1_1,probe_mut2vec,17445,300,float32,embeddings,data/embeddings/probe_mut2vec/embeddings.npz,data/embeddings/probe_mut2vec/genes.tsv,data/embeddings/probe_mut2vec/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_mut2vec_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,MUT2VEC,PROBE authors / original PROBE human representation archive,,,,10844260 +probe_pfam_v1_1,probe_pfam,18873,6227,float32,embeddings,data/embeddings/probe_pfam/embeddings.npz,data/embeddings/probe_pfam/genes.tsv,data/embeddings/probe_pfam/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_pfam_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,PFAM,PROBE authors / original PROBE human representation archive,,,,548968 +probe_protvec_v1_1,probe_protvec,18873,100,float32,embeddings,data/embeddings/probe_protvec/embeddings.npz,data/embeddings/probe_protvec/genes.tsv,data/embeddings/probe_protvec/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_protvec_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,PROTVEC,PROBE authors / original PROBE human representation archive,,,,7066137 +probe_seqvec_v1_1,probe_seqvec,18873,1024,float32,embeddings,data/embeddings/probe_seqvec/embeddings.npz,data/embeddings/probe_seqvec/genes.tsv,data/embeddings/probe_seqvec/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_seqvec_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,SEQVEC,PROBE authors / original PROBE human representation archive,,,,71841293 +probe_t5_v1_1,probe_t5,18453,1024,float32,embeddings,data/embeddings/probe_t5/embeddings.npz,data/embeddings/probe_t5/genes.tsv,data/embeddings/probe_t5/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_t5_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,T5,PROBE authors / original PROBE human representation archive,,,,70153727 +probe_tcga_embedding_v1_1,probe_tcga_embedding,18221,50,float32,embeddings,data/embeddings/probe_tcga_embedding/embeddings.npz,data/embeddings/probe_tcga_embedding/genes.tsv,data/embeddings/probe_tcga_embedding/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_tcga_embedding_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,TCGA-EMBEDDING,PROBE authors / original PROBE human representation archive,,,,3378621 +probe_unirep_v1_1,probe_unirep,18873,5700,float32,embeddings,data/embeddings/probe_unirep/embeddings.npz,data/embeddings/probe_unirep/genes.tsv,data/embeddings/probe_unirep/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_unirep_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,UNIREP,PROBE authors / original PROBE human representation archive,,,,410932244 +probe_xlnet_v1_1,probe_xlnet,18453,1024,float32,embeddings,data/embeddings/probe_xlnet/embeddings.npz,data/embeddings/probe_xlnet/genes.tsv,data/embeddings/probe_xlnet/metadata.json,,protein sequence / protein representation from PROBE human UniProt archive,probe_xlnet_v1_1,Official PROBE human_representation_vectors.tar.gz archive from Zenodo record 5795850 plus official HUBioDataLab/PROBE repository,XLNET,PROBE authors / original PROBE human representation archive,,,,70519787 +pubmedbert_gene_protein_summary_v1_1,pubmedbert_gene_protein_summary,19168,768,float32,embeddings,data/embeddings/pubmedbert_gene_protein_summary/embeddings.npz,data/embeddings/pubmedbert_gene_protein_summary/genes.tsv,data/embeddings/pubmedbert_gene_protein_summary/metadata.json,,literature_text_gene_summary,PubMedBERT-GeneProteinSummary-v1.1,GenePT Zenodo v2 summaries,,Daniel locally,microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext,"final hidden layer mean pooling over valid non-special, non-padding tokens",,54786752 +rnafm_full_transcript_v1_1,rnafm_full_transcript,19976,640,float32,embeddings,data/embeddings/rnafm_full_transcript/embeddings.npz,data/embeddings/rnafm_full_transcript/genes.tsv,data/embeddings/rnafm_full_transcript/metadata.json,processed_qc_passed,,rnafm_full_transcript_v1_1,,,,/local/home/dtabares/data/raw_embeddings/rnafm_full_transcript_v1_1/hf_model/RNA-FM_pretrained.pth,Final layer 12 mean token pooling over real RNA tokens only; BOS/EOS excluded; long transcripts split into chunks and chunk means length-weighted,,47521441 +scfoundation_gene_pos_v1_1,scfoundation_gene_pos,19062,768,float32,embeddings,data/embeddings/scfoundation_gene_pos/embeddings.npz,data/embeddings/scfoundation_gene_pos/genes.tsv,data/embeddings/scfoundation_gene_pos/metadata.json,,single_cell_expression_foundation_model / learned gene identity-position embedding,scFoundation-gene-pos-v1.1,Official scFoundation GitHub repository and official SharePoint checkpoint linked from the repository,scFoundation / xTrimoGene,Extracted locally from official scFoundation gene sub-checkpoint model.pos_emb.weight,,,,54234742 +scgpt_pancancer_v1_1,scgpt_pancancer,19071,512,float32,embeddings,data/embeddings/scgpt_pancancer/embeddings.npz,data/embeddings/scgpt_pancancer/genes.tsv,data/embeddings/scgpt_pancancer/metadata.json,processed_qc_passed,,scgpt_pancancer_v1_1,,,,,,,36187165 +scgpt_whole_human_v1_1,scgpt_whole_human,19280,512,float32,embeddings,data/embeddings/scgpt_whole_human/embeddings.npz,data/embeddings/scgpt_whole_human/genes.tsv,data/embeddings/scgpt_whole_human/metadata.json,,single_cell_expression_foundation_model,scGPT-whole-human-v1.1,Official scGPT whole-human Google Drive checkpoint linked from the scGPT project,,Extracted locally from official scGPT whole-human model input gene-token embedding layer,,,,36573474 +scprint_medium_v1_5_gene_v1_1,scprint_medium_v1_5_gene,19995,256,float32,embeddings,data/embeddings/scprint_medium_v1_5_gene/embeddings.npz,data/embeddings/scprint_medium_v1_5_gene/genes.tsv,data/embeddings/scprint_medium_v1_5_gene/metadata.json,,single_cell_expression_foundation_model / learned gene-token embedding,scPRINT-medium-v1.5-gene-v1.1,Official scPRINT HuggingFace repository jkobject/scPRINT,scPRINT,Extracted locally from official scPRINT medium-v1.5 checkpoint,,,,18989317 +spliceai_canonical_splice_site_profile_v1_1,spliceai_canonical_splice_site_profile,20097,606,float32,embeddings,data/embeddings/spliceai_canonical_splice_site_profile/embeddings.npz,data/embeddings/spliceai_canonical_splice_site_profile/genes.tsv,data/embeddings/spliceai_canonical_splice_site_profile/metadata.json,processed_qc_passed,splicing sequence model / canonical transcript splice-site profiles,SpliceAI-canonical-splice-site-profile-zscore-v1.1,,,,,,,41717181 +uce_33l_v1_1,uce_33l,19082,1280,float32,embeddings,data/embeddings/uce_33l/embeddings.npz,data/embeddings/uce_33l/genes.tsv,data/embeddings/uce_33l/metadata.json,,single_cell_expression_foundation_model / protein-informed gene-token embedding,UCE-33L-v1.1,Official UCE GitHub repository plus official Figshare model files,Universal Cell Embeddings (UCE),Extracted locally from official UCE all_tokens and 33-layer checkpoint gene encoder,,,,89725640 diff --git a/data/manifest.parquet b/data/manifest.parquet index 9d3fda145dc40d195ba161cda350b3bc328c3480..9694ac668d1740ffda6e121dd871c868e319900d 100644 --- a/data/manifest.parquet +++ b/data/manifest.parquet @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:250fc8907a48d28db43fac9840507d0d0a095f549add57c7371492366cdddcf2 -size 26881 +oid sha256:6debe725d13850d0059a00c79f5fb59d9719e7299b12e385cb71f1094264adca +size 27316 diff --git a/data/master_gene_table_v1_1_enriched.csv b/data/master_gene_table.csv similarity index 100% rename from data/master_gene_table_v1_1_enriched.csv rename to data/master_gene_table.csv diff --git a/data/master_gene_table_v1_1_enriched.parquet b/data/master_gene_table.parquet similarity index 100% rename from data/master_gene_table_v1_1_enriched.parquet rename to data/master_gene_table.parquet diff --git a/docs/per_embedding_readme_manual_todo.csv b/docs/per_embedding_readme_manual_todo.csv new file mode 100644 index 0000000000000000000000000000000000000000..7fb46afec53f8be69568749ed1607806201f6a46 --- /dev/null +++ b/docs/per_embedding_readme_manual_todo.csv @@ -0,0 +1,69 @@ +embedding_id,readme_path,unknown_genes_tsv_columns,manual_processing_notes_status,former_processing_embedding_id +alphagenome_tss1mb_regulatory_tracksummary_pca256,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/README.md,,TODO,alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1 +basenji2_tss_131kb_human_central3,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/basenji2_tss_131kb_human_central3/README.md,,TODO,basenji2_tss_131kb_human_central3_v1_1 +biobert_gene_protein_summary,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/biobert_gene_protein_summary/README.md,,TODO,biobert_gene_protein_summary_v1_1 +bioconceptvec_cbow,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/bioconceptvec_cbow/README.md,,TODO,bioconceptvec_cbow_v1_1 +bioconceptvec_fasttext,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/bioconceptvec_fasttext/README.md,,TODO,bioconceptvec_fasttext_v1_1 +bioconceptvec_glove,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/bioconceptvec_glove/README.md,,TODO,bioconceptvec_glove_v1_1 +bioconceptvec_skipgram,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/bioconceptvec_skipgram/README.md,,TODO,bioconceptvec_skipgram_v1_1 +cmap_gse106127_cgs_mean_landmark978_zscore,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/README.md,n_cgs_signatures;n_cell_lines,TODO,cmap_gse106127_cgs_mean_landmark978_zscore_v1_1 +deepnf_string_v12_ppmi_svd_fusion,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/README.md,,TODO,deepnf_string_v12_ppmi_svd_fusion_v1_1 +deepsea_tss_1kb_epigenomic_probs,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/deepsea_tss_1kb_epigenomic_probs/README.md,,TODO,deepsea_tss_1kb_epigenomic_probs_v1_1 +depmap_crispr_gene_effect_pca256,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/depmap_crispr_gene_effect_pca256/README.md,source_column,TODO,depmap_crispr_gene_effect_pca256_v1_1 +dnabert2_117m_cds,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/dnabert2_117m_cds/README.md,dnabert2_sequence_length,TODO,dnabert2_117m_cds_v1_1 +enformer_tss_393kb_human_central3,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/enformer_tss_393kb_human_central3/README.md,,TODO,enformer_tss_393kb_human_central3_v1_1 +esm2,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/esm2/README.md,,CURATED,esm2_v1_1 +frogs_archs4_256,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/frogs_archs4_256/README.md,,TODO,frogs_archs4_256_v1_1 +gene2vec,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/gene2vec/README.md,,TODO,gene2vec_v1_1 +genecompass_base_gene_token,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/genecompass_base_gene_token/README.md,genecompass_token_id,TODO,genecompass_base_gene_token_v1_1 +geneformer_v2_316m,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/geneformer_v2_316m/README.md,,TODO,geneformer_v2_316m_v1_1 +genept_ada_conservative,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/genept_ada_conservative/README.md,,TODO,genept_ada_v1_1_conservative +genept_model3_gene_protein_conservative,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/genept_model3_gene_protein_conservative/README.md,,TODO,genept_model3_gene_protein_v1_1_conservative +gtex_tissue_median_log1p_zscore,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/gtex_tissue_median_log1p_zscore/README.md,,TODO,gtex_tissue_median_log1p_zscore_v1_1 +helix_mrna_full_transcript,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/helix_mrna_full_transcript/README.md,cds_start_0based;cds_end_0based_exclusive;helix_input_length;helix_token_count_used_for_mean_pooling;helix_chunk_count,TODO,helix_mrna_full_transcript_v1_1 +hetionet_nonppi_biomedical_context_tfidf_svd256,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/README.md,,TODO,hetionet_nonppi_biomedical_context_tfidf_svd256_v1_1 +hpa_normal_ihc_tissue_level_zscore,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/hpa_normal_ihc_tissue_level_zscore/README.md,hpa_gene;hpa_gene_name,TODO,hpa_normal_ihc_tissue_level_zscore_v1_1 +hpa_subcellular_location_multihot_zscore,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/hpa_subcellular_location_multihot_zscore/README.md,hpa_gene;hpa_gene_name;hpa_reliability,TODO,hpa_subcellular_location_multihot_zscore_v1_1 +hyenadna_medium_160k_cds,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/hyenadna_medium_160k_cds/README.md,hyenadna_token_count_used_for_mean_pooling;hyenadna_chunk_count,TODO,hyenadna_medium_160k_cds_v1_1 +mahi_all_contexts_mean,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/mahi_all_contexts_mean/README.md,mahi_context_count,TODO,mahi_all_contexts_mean_v1_1 +mahi_global,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/mahi_global/README.md,mahi_context_count,TODO,mahi_global_v1_1 +mashup_string,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/mashup_string/README.md,,TODO,mashup_string_v1_1 +msa_transformer_depth1_uniprot_human,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/msa_transformer_depth1_uniprot_human/README.md,,TODO,msa_transformer_depth1_uniprot_human_v1_1 +newt_archs4_256,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/newt_archs4_256/README.md,,TODO,newt_archs4_256_v1_1 +newt_cellnet,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/newt_cellnet/README.md,,TODO,newt_cellnet_v1_1 +newt_go_256,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/newt_go_256/README.md,,TODO,newt_go_256_v1_1 +newt_go_graph,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/newt_go_graph/README.md,,TODO,newt_go_graph_v1_1 +newt_msigdb_bundle,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/newt_msigdb_bundle/README.md,,TODO,newt_msigdb_bundle_v1_1 +node2vec_consensus_ppi,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/node2vec_consensus_ppi/README.md,,TODO,node2vec_consensus_ppi_v1_1 +nt_v2_500m_multispecies_cds,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/nt_v2_500m_multispecies_cds/README.md,nt_v2_token_count_used_for_mean_pooling;effective_chunk_nt,TODO,nt_v2_500m_multispecies_cds_v1_1 +opa2vec_goa_human,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/opa2vec_goa_human/README.md,,TODO,opa2vec_goa_human_v1_1 +orthrus_base_4track_full_transcript,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/orthrus_base_4track_full_transcript/README.md,orthrus_sequence_length,TODO,orthrus_base_4track_full_transcript_v1_1 +primekg_nonppi_biomedical_context_tfidf_svd256,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/README.md,,TODO,primekg_nonppi_biomedical_context_tfidf_svd256_v1_1 +probe_aac,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_aac/README.md,,TODO,probe_aac_v1_1 +probe_albert,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_albert/README.md,,TODO,probe_albert_v1_1 +probe_apaac,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_apaac/README.md,,TODO,probe_apaac_v1_1 +probe_bert_bfd,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_bert_bfd/README.md,,TODO,probe_bert_bfd_v1_1 +probe_bert_pfam,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_bert_pfam/README.md,,TODO,probe_bert_pfam_v1_1 +probe_blast,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_blast/README.md,,TODO,probe_blast_v1_1 +probe_cpc_prot,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_cpc_prot/README.md,,TODO,probe_cpc_prot_v1_1 +probe_esmb1,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_esmb1/README.md,,TODO,probe_esmb1_v1_1 +probe_gene2vec_uniprot,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_gene2vec_uniprot/README.md,,TODO,probe_gene2vec_uniprot_v1_1 +probe_hmmer,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_hmmer/README.md,,TODO,probe_hmmer_v1_1 +probe_ksep,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_ksep/README.md,,TODO,probe_ksep_v1_1 +probe_learned_vec,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_learned_vec/README.md,,TODO,probe_learned_vec_v1_1 +probe_mut2vec,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_mut2vec/README.md,,TODO,probe_mut2vec_v1_1 +probe_pfam,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_pfam/README.md,,TODO,probe_pfam_v1_1 +probe_protvec,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_protvec/README.md,,TODO,probe_protvec_v1_1 +probe_seqvec,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_seqvec/README.md,,TODO,probe_seqvec_v1_1 +probe_t5,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_t5/README.md,,TODO,probe_t5_v1_1 +probe_tcga_embedding,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_tcga_embedding/README.md,,TODO,probe_tcga_embedding_v1_1 +probe_unirep,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_unirep/README.md,,TODO,probe_unirep_v1_1 +probe_xlnet,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/probe_xlnet/README.md,,TODO,probe_xlnet_v1_1 +pubmedbert_gene_protein_summary,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/pubmedbert_gene_protein_summary/README.md,,TODO,pubmedbert_gene_protein_summary_v1_1 +rnafm_full_transcript,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/rnafm_full_transcript/README.md,rnafm_sequence_length,TODO,rnafm_full_transcript_v1_1 +scfoundation_gene_pos,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/scfoundation_gene_pos/README.md,source_checkpoint_indices_examples,TODO,scfoundation_gene_pos_v1_1 +scgpt_pancancer,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/scgpt_pancancer/README.md,,TODO,scgpt_pancancer_v1_1 +scgpt_whole_human,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/scgpt_whole_human/README.md,,TODO,scgpt_whole_human_v1_1 +scprint_medium_v1_5_gene,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/scprint_medium_v1_5_gene/README.md,,TODO,scprint_medium_v1_5_gene_v1_1 +spliceai_canonical_splice_site_profile,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/spliceai_canonical_splice_site_profile/README.md,spliceai_site_status;n_splice_sites;n_donor_sites;n_acceptor_sites,TODO,spliceai_canonical_splice_site_profile_v1_1 +uce_33l,/local/home/dtabares/data/hf_gene_embeddings/data/embeddings/uce_33l/README.md,source_chromosomes_examples;source_starts_examples,TODO,uce_33l_v1_1