Clean public folder structure and master table filenames

.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

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/<embedding_id>/<embedding_id>_embeddings.npz`
 - `data/embeddings/<embedding_id>/<embedding_id>_genes.tsv`
 - `data/embeddings/<embedding_id>/<embedding_id>_metadata.json`

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

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<br>1<br>2<br>3 | Zero-based row index in this embedding matrix and genes.tsv file. |
+| `ensembl_gene_id` | 20084 | 20084 | ENSG00000000003<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 20084 | 20057 | TSPAN6<br>TNMD<br>DPM1<br>SCYL3 | Gene symbol associated with the Ensembl gene ID. |
+| `gene_type` | 20084 | 1 | protein_coding | Gene biotype, usually protein_coding. |
+| `chromosome` | 20084 | 24 | chrX<br>chr20<br>chr1<br>chr6 | Chromosome for the gene or modeled interval. |
+| `gene_start` | 20084 | 20026 | 100627108<br>100584936<br>50934852<br>169846981 | Gene start coordinate in the stated genome assembly. |
+| `gene_end` | 20084 | 19975 | 100639991<br>100599885<br>50959140<br>169894267 | Gene end coordinate in the stated genome assembly. |
+| `strand` | 20084 | 2 | -<br>+ | Genomic strand. |
+| `assembly` | 20084 | 1 | GRCh38 | Genome assembly used for coordinates. |
+| `tss_1based` | 20084 | 20034 | 100639991<br>100584936<br>50959140<br>169894267 | Transcription start site coordinate in 1-based coordinates. |
+| `input_interval` | 20084 | 20050 | chrX:100115702-101164278:-<br>chrX:100060648-101109224:+<br>chr20:50434851-51483427:-<br>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.

data/embeddings/{alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_metadata.json → 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"
-}
+  "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"
+  }
+}

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<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 20097 | 20070 | TSPAN6<br>TNMD<br>DPM1<br>SCYL3 | Gene symbol associated with the Ensembl gene ID. |
+| `entrez_id` | 19545 | 19511 | 7105<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 20045 | 19860 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `gene_type` | 20097 | 1 | protein_coding | Gene biotype, usually protein_coding. |
+| `chromosome` | 20097 | 25 | chrX<br>chr20<br>chr1<br>chr6 | Chromosome for the gene or modeled interval. |
+| `gene_start` | 20097 | 20039 | 100627108<br>100584936<br>50934852<br>169846981 | Gene start coordinate in the stated genome assembly. |
+| `gene_end` | 20097 | 19988 | 100639991<br>100599885<br>50959140<br>169894267 | Gene end coordinate in the stated genome assembly. |
+| `strand` | 20097 | 2 | -<br>+ | Genomic strand. |
+| `tss_1based` | 20097 | 20047 | 100639991<br>100584936<br>50959140<br>169894267 | Transcription start site coordinate in 1-based coordinates. |
+| `tss_0based` | 20097 | 20047 | 100639990<br>100584935<br>50959139<br>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.

data/embeddings/{basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_metadata.json → 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"
-}
+  "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"
+  }
+}

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<br>1<br>2<br>3 | Zero-based row index in this embedding matrix and genes.tsv file. |
+| `ensembl_gene_id` | 19168 | 19168 | ENSG00000000003<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 19168 | 19143 | TSPAN6<br>TNMD<br>DPM1<br>SCYL3 | Gene symbol associated with the Ensembl gene ID. |
+| `hgnc_approved_symbol` | 19082 | 19083 | TSPAN6<br>TNMD<br>DPM1<br>SCYL3 | HGNC-approved gene symbol from the master table. |
+| `entrez_id` | 19118 | 19092 | 7105<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 19153 | 18980 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `summary_key` | 19168 | 19082 | TSPAN6<br>TNMD<br>DPM1<br>SCYL3 | Identifier used to retrieve the gene/protein text summary. |
+| `summary_mapping_source` | 19168 | 3 | gene_symbol<br>map_symbols_all_fallback<br>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.

data/embeddings/{biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_metadata.json → biobert_gene_protein_summary/metadata.json}

@@ -23,5 +23,13 @@
   "zero_vectors": 0,
   "nan_count": 0,
   "inf_count": 0,
-  "runtime_minutes": 3.477354129155477
-}
+  "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"
+  }
+}

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<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 17040 | 17040 | TSPAN6<br>TNMD<br>DPM1<br>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<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 17039 | 16973 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `canonical_transcript_id` | 17040 | 17040 | ENST00000373020<br>ENST00000373031<br>ENST00000371588<br>ENST00000367771 | Canonical transcript identifier recorded or used for this gene. |
+| `canonical_protein_id` | 17040 | 17040 | ENSP00000362111<br>ENSP00000362122<br>ENSP00000360644<br>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<br>2 | Number of source keys mapped or collapsed to this final gene. |
+| `source_keys_examples` | 17040 | 17040 | Gene_7105<br>Gene_64102<br>Gene_8813<br>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.

data/embeddings/{bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_metadata.json → 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"
-}
+  "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"
+  }
+}

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<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 17040 | 17040 | TSPAN6<br>TNMD<br>DPM1<br>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<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 17039 | 16973 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `canonical_transcript_id` | 17040 | 17040 | ENST00000373020<br>ENST00000373031<br>ENST00000371588<br>ENST00000367771 | Canonical transcript identifier recorded or used for this gene. |
+| `canonical_protein_id` | 17040 | 17040 | ENSP00000362111<br>ENSP00000362122<br>ENSP00000360644<br>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<br>2 | Number of source keys mapped or collapsed to this final gene. |
+| `source_keys_examples` | 17040 | 17040 | Gene_7105<br>Gene_64102<br>Gene_8813<br>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.

data/embeddings/{bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_metadata.json → 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"
-}
+  "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"
+  }
+}

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<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 17040 | 17040 | TSPAN6<br>TNMD<br>DPM1<br>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<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 17039 | 16973 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `canonical_transcript_id` | 17040 | 17040 | ENST00000373020<br>ENST00000373031<br>ENST00000371588<br>ENST00000367771 | Canonical transcript identifier recorded or used for this gene. |
+| `canonical_protein_id` | 17040 | 17040 | ENSP00000362111<br>ENSP00000362122<br>ENSP00000360644<br>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<br>2 | Number of source keys mapped or collapsed to this final gene. |
+| `source_keys_examples` | 17040 | 17040 | Gene_7105<br>Gene_64102<br>Gene_8813<br>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.

data/embeddings/{bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_metadata.json → 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"
-}
+  "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"
+  }
+}

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<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 17040 | 17040 | TSPAN6<br>TNMD<br>DPM1<br>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<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 17039 | 16973 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `canonical_transcript_id` | 17040 | 17040 | ENST00000373020<br>ENST00000373031<br>ENST00000371588<br>ENST00000367771 | Canonical transcript identifier recorded or used for this gene. |
+| `canonical_protein_id` | 17040 | 17040 | ENSP00000362111<br>ENSP00000362122<br>ENSP00000360644<br>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<br>2 | Number of source keys mapped or collapsed to this final gene. |
+| `source_keys_examples` | 17040 | 17040 | Gene_7105<br>Gene_64102<br>Gene_8813<br>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.

data/embeddings/{bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_metadata.json → 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"
-}
+  "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"
+  }
+}

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<br>1<br>2<br>3 | Zero-based row index in this embedding matrix and genes.tsv file. |
+| `ensembl_gene_id` | 4070 | 4070 | ENSG00000000003<br>ENSG00000000457<br>ENSG00000000938<br>ENSG00000001084 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 4070 | 4063 | TSPAN6<br>SCYL3<br>FGR<br>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<br>57147<br>2268<br>2729 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 4070 | 4060 | O43657<br>Q8IZE3<br>P09769<br>P48506 | UniProt accession or identifier, when available. |
+| `source_gene_symbol` | 4070 | 4063 | TSPAN6<br>SCYL3<br>FGR<br>GCLC | Original gene symbol from the source before final mapping. |
+| `n_cgs_signatures` | 4070 | 6 | 9<br>1<br>8<br>7 | TODO: define this source-specific column during manual curation. |
+| `n_cell_lines` | 4070 | 6 | 9<br>1<br>8<br>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.

data/embeddings/{cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_metadata.json → 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"
   }
-}
+}

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<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 19071 | 19071 | TSPAN6<br>TNMD<br>DPM1<br>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<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 19060 | 19010 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `canonical_transcript_id` | 19071 | 19071 | ENST00000373020<br>ENST00000373031<br>ENST00000371588<br>ENST00000367771 | Canonical transcript identifier recorded or used for this gene. |
+| `canonical_protein_id` | 19071 | 19071 | ENSP00000362111<br>ENSP00000362122<br>ENSP00000360644<br>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<br>2<br>3 | Number of source keys mapped or collapsed to this final gene. |
+| `source_keys_examples` | 19071 | 19071 | 9606.ENSP00000362111<br>9606.ENSP00000362122<br>9606.ENSP00000360638<br>9606.ENSP00000356746 | Example source keys mapped to this final gene. |
+| `source_row_indices_examples` | 19071 | 19071 | 10924<br>10928<br>10628<br>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<br>ensembl_protein_id_exact;string_alias_entrez;string_id_exact<br>ensembl_protein_id_exact;string_alias_entrez;string_id_exact;ensembl_protein_id_exact;string_alias_entrez;string_id_exact;string_preferred_symbol<br>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.

data/embeddings/{deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_metadata.json → 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"
   }
-}
+}

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<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 20097 | 20070 | TSPAN6<br>TNMD<br>DPM1<br>SCYL3 | Gene symbol associated with the Ensembl gene ID. |
+| `entrez_id` | 19545 | 19511 | 7105<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 20045 | 19860 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `gene_type` | 20097 | 1 | protein_coding | Gene biotype, usually protein_coding. |
+| `chromosome` | 20097 | 25 | chrX<br>chr20<br>chr1<br>chr6 | Chromosome for the gene or modeled interval. |
+| `gene_start` | 20097 | 20039 | 100627108<br>100584936<br>50934852<br>169846981 | Gene start coordinate in the stated genome assembly. |
+| `gene_end` | 20097 | 19988 | 100639991<br>100599885<br>50959140<br>169894267 | Gene end coordinate in the stated genome assembly. |
+| `strand` | 20097 | 2 | -<br>+ | Genomic strand. |
+| `tss_1based` | 20097 | 20047 | 100639991<br>100584936<br>50959140<br>169894267 | Transcription start site coordinate in 1-based coordinates. |
+| `bed_start_0based` | 20097 | 20047 | 100639490<br>100584435<br>50958639<br>169893766 | BED-format interval start coordinate, 0-based. |
+| `bed_end_0based` | 20097 | 20047 | 100640490<br>100585435<br>50959639<br>169894766 | BED-format interval end coordinate, half-open. |
+| `bed_name` | 20097 | 20097 | ENSG00000000003|TSPAN6<br>ENSG00000000005|TNMD<br>ENSG00000000419|DPM1<br>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.

data/embeddings/{deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_metadata.json → 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"
-}
+  "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"
+  }
+}

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<br>1<br>2<br>3 | Zero-based row index in this embedding matrix and genes.tsv file. |
+| `ensembl_gene_id` | 18562 | 18562 | ENSG00000000003<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `gene_symbol` | 18562 | 18537 | TSPAN6<br>TNMD<br>DPM1<br>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<br>64102<br>8813<br>57147 | NCBI Entrez Gene identifier, when available. |
+| `uniprot_id` | 18562 | 18498 | O43657<br>Q9H2S6<br>O60762<br>Q8IZE3 | UniProt accession or identifier, when available. |
+| `source_entrez_id` | 18562 | 18529 | 7105<br>64102<br>8813<br>57147 | Entrez Gene identifier from the original source before final mapping. |
+| `source_gene_symbol` | 18562 | 18529 | TSPAN6<br>TNMD<br>DPM1<br>SCYL3 | Original gene symbol from the source before final mapping. |
+| `source_column` | 18562 | 18529 | TSPAN6 (7105)<br>TNMD (64102)<br>DPM1 (8813)<br>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.

data/embeddings/{depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_metadata.json → 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"
   }
-}
+}

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<br>ENSG00000000005<br>ENSG00000000419<br>ENSG00000000457 | Ensembl gene identifier; main stable identifier used to align genes across embeddings. |
+| `transcript_id` | 20085 | 20085 | ENST00000373020<br>ENST00000373031<br>ENST00000371588<br>ENST00000367771 | Transcript identifier used for sequence-based processing. |
+| `dnabert2_sequence_length` | 20085 | 2255 | 738<br>954<br>783<br>2067 | TODO: define this source-specific column during manual curation. |
+| `num_chunks` | 20085 | 19 | 1<br>2<br>3<br>4 | Number of chunks/windows used when the input was too long. |
+| `was_chunked` | 20085 | 2 | False<br>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.