danieltslz commited on
Commit
39a01a9
·
verified ·
1 Parent(s): 10a4b00

Initial private staging upload of 68 standardized human gene embeddings

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. README.md +38 -0
  2. data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_embeddings.npz +3 -0
  3. data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_genes.tsv +0 -0
  4. data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_metadata.json +34 -0
  5. data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_embeddings.npz +3 -0
  6. data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_genes.tsv +0 -0
  7. data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_metadata.json +35 -0
  8. data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_embeddings.npz +3 -0
  9. data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_genes.tsv +0 -0
  10. data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_metadata.json +27 -0
  11. data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_embeddings.npz +3 -0
  12. data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_genes.tsv +0 -0
  13. data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_metadata.json +31 -0
  14. data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_embeddings.npz +3 -0
  15. data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_genes.tsv +0 -0
  16. data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_metadata.json +31 -0
  17. data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_embeddings.npz +3 -0
  18. data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_genes.tsv +0 -0
  19. data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_metadata.json +31 -0
  20. data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_embeddings.npz +3 -0
  21. data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_genes.tsv +0 -0
  22. data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_metadata.json +31 -0
  23. data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_embeddings.npz +3 -0
  24. data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_genes.tsv +0 -0
  25. data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_metadata.json +38 -0
  26. data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_embeddings.npz +3 -0
  27. data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_genes.tsv +0 -0
  28. data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_metadata.json +148 -0
  29. data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_embeddings.npz +3 -0
  30. data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_genes.tsv +0 -0
  31. data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_metadata.json +30 -0
  32. data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_embeddings.npz +3 -0
  33. data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_genes.tsv +0 -0
  34. data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_metadata.json +36 -0
  35. data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_embeddings.npz +3 -0
  36. data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_genes.tsv +0 -0
  37. data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_metadata.json +35 -0
  38. data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_embeddings.npz +3 -0
  39. data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_genes.tsv +0 -0
  40. data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_metadata.json +36 -0
  41. data/embeddings/esm2_v1_1/esm2_v1_1_embeddings.npz +3 -0
  42. data/embeddings/esm2_v1_1/esm2_v1_1_genes.tsv +0 -0
  43. data/embeddings/esm2_v1_1/esm2_v1_1_metadata.json +37 -0
  44. data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_embeddings.npz +3 -0
  45. data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_genes.tsv +0 -0
  46. data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_metadata.json +34 -0
  47. data/embeddings/gene2vec_v1_1/gene2vec_v1_1_embeddings.npz +3 -0
  48. data/embeddings/gene2vec_v1_1/gene2vec_v1_1_genes.tsv +0 -0
  49. data/embeddings/gene2vec_v1_1/gene2vec_v1_1_metadata.json +30 -0
  50. data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_embeddings.npz +3 -0
README.md ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ pretty_name: Unified human gene embeddings
4
+ tags:
5
+ - biology
6
+ - genomics
7
+ - transcriptomics
8
+ - gene-embeddings
9
+ - computational-biology
10
+ ---
11
+
12
+ # Unified human gene embeddings
13
+
14
+ Private staging repository for Daniel's standardized human gene embedding resource.
15
+
16
+ ## Status
17
+
18
+ This repository is private staging. Do not make it public until redistribution permissions and licenses have been checked source by source.
19
+
20
+ ## Contents
21
+
22
+ - `data/manifest.csv`
23
+ - `data/manifest.parquet`
24
+ - `data/master_gene_table_v1_1_enriched.csv`
25
+ - `data/master_gene_table_v1_1_enriched.parquet`
26
+ - `data/embeddings/<embedding_id>/<embedding_id>_embeddings.npz`
27
+ - `data/embeddings/<embedding_id>/<embedding_id>_genes.tsv`
28
+ - `data/embeddings/<embedding_id>/<embedding_id>_metadata.json`
29
+ - `docs/processed_embeddings_README.md`
30
+ - `scripts/load_embedding.py`
31
+
32
+ ## Current staged embeddings
33
+
34
+ Number of staged embeddings: 68
35
+
36
+ ## Important note
37
+
38
+ AlphaGenome-derived outputs require separate review of output use and redistribution conditions.
data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ea0f71bb0966c63ef59cc7477f0218abef0accfab4d31279a3c45bd4b9a8ca4
3
+ size 19270344
data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1/alphagenome_tss1mb_regulatory_tracksummary_pca256_v1_1_metadata.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "AlphaGenome-TSS1Mb-regulatoryTrackSummary-PCA256-v1.1",
3
+ "status": "processed_qc_passed",
4
+ "raw_source_directory": "/local/home/dtabares/data/raw_embeddings/alphagenome_tss1mb_regulatory_v1_1",
5
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
6
+ "n_chunk_files": 204,
7
+ "n_raw_rows": 20084,
8
+ "n_unique_raw_genes": 20084,
9
+ "n_target_genes": 20084,
10
+ "n_covered_genes": 20084,
11
+ "n_missing_genes": 0,
12
+ "n_raw_summary_features": 1250,
13
+ "n_pca_components": 256,
14
+ "raw_outputs_saved": false,
15
+ "input_window": "1 Mb centered on canonical TSS from master table",
16
+ "coordinate_system": "GRCh38; master table 1-based coordinates converted to AlphaGenome 0-based half-open interval anchor",
17
+ "requested_outputs": [
18
+ "CAGE",
19
+ "RNA_SEQ",
20
+ "DNASE",
21
+ "ATAC",
22
+ "CHIP_HISTONE",
23
+ "CHIP_TF",
24
+ "PROCAP"
25
+ ],
26
+ "summary_statistics": [
27
+ "central_1kb_mean",
28
+ "central_10kb_mean",
29
+ "central_100kb_mean",
30
+ "full_1mb_mean",
31
+ "central_10kb_max"
32
+ ],
33
+ "transform": "signed log1p per summary feature, z-score across genes, PCA to 256 dimensions"
34
+ }
data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b6c7615d0441ad3380293b7a7cb6da5d534d26050ff70b0493cb0aba07eddbe
3
+ size 389641898
data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/basenji2_tss_131kb_human_central3_v1_1/basenji2_tss_131kb_human_central3_v1_1_metadata.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "Basenji2-TSS-131kb-human-central3-log1p-zscore-v1.1",
3
+ "status": "processed_qc_passed",
4
+ "modality": "regulatory DNA / Basenji2 TSS sequence model",
5
+ "model": "Basenji2 PyTorch port",
6
+ "weights": "/local/home/dtabares/.cache/pooch/c5a450027ee5042f4208cde76d593de8-basenji2.pth",
7
+ "input_tensor_layout": "channels_first, shape = batch x 4 x 131072",
8
+ "output_shape_per_sequence": [
9
+ 896,
10
+ 5313
11
+ ],
12
+ "pooled_representation": "mean of central 3 output bins around TSS",
13
+ "raw_output_file": "raw_basenji2_human_central3_mean_tracks.npy",
14
+ "transformation": "log1p raw human central3 mean tracks, then dimension-wise z-score across genes",
15
+ "final_embedding_file": "embeddings.npy",
16
+ "input_sequence_type": "GRCh38 genomic DNA, 131072 bp centered on gene TSS",
17
+ "input_sequence_source": "/local/home/dtabares/data/raw_embeddings/deepsea_tss_1kb_v1_1/reference_genome/GRCh38.primary_assembly.genome.fa",
18
+ "annotation_source": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
19
+ "genome_build": "GRCh38",
20
+ "sequence_length_bp": 131072,
21
+ "target_length_bins": 896,
22
+ "central_bins_used_0based": [
23
+ 447,
24
+ 448,
25
+ 449
26
+ ],
27
+ "tss_rule": "plus strand: gene_start; minus strand: gene_end; coordinates from master_gene_table_v1_1_enriched.csv",
28
+ "strand_handling": "genomic reference orientation; no gene-strand reverse-complementing",
29
+ "boundary_handling": "N-padding if the 131072 bp window extends beyond chromosome boundaries",
30
+ "n_master_rows_input": 20097,
31
+ "n_valid_tss_windows": 20097,
32
+ "n_skipped_tss_windows": 0,
33
+ "dimension": 5313,
34
+ "created_at": "2026-07-10T06:17:21"
35
+ }
data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8b14297862ad561278b051579ad2ee2309083d85498ab04260662e3efd0c6eb
3
+ size 54606212
data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/biobert_gene_protein_summary_v1_1/biobert_gene_protein_summary_v1_1_metadata.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "BioBERT-GeneProteinSummary-v1.1",
3
+ "modality": "literature_text_gene_summary",
4
+ "download_source": "GenePT Zenodo v2 summaries",
5
+ "original_text_source": "NCBI gene card summaries plus UniProt protein summaries when available",
6
+ "embedding_generated_by": "Daniel locally",
7
+ "model_checkpoint": "dmis-lab/biobert-base-cased-v1.2",
8
+ "pooling_strategy": "final hidden layer mean pooling over valid non-special, non-padding tokens",
9
+ "summary_json": "/local/home/dtabares/data/raw_embeddings/genept_text_summaries_v1_1/GenePT_emebdding_v2/NCBI_UniProt_summary_of_genes.json",
10
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
11
+ "master_rows": 20097,
12
+ "mapped_genes": 19168,
13
+ "master_coverage": 0.9537741951535055,
14
+ "embedding_dim": 768,
15
+ "matrix_shape": [
16
+ 19168,
17
+ 768
18
+ ],
19
+ "dtype": "float32",
20
+ "batch_size": 16,
21
+ "max_length": 512,
22
+ "device": "cuda",
23
+ "zero_vectors": 0,
24
+ "nan_count": 0,
25
+ "inf_count": 0,
26
+ "runtime_minutes": 3.477354129155477
27
+ }
data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55234e84e1299be358d35d3dd02d4664f6f531513af974f39a642fae2f83577f
3
+ size 6455585
data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_metadata.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "BioConceptVec-CBOW-v1.1",
3
+ "modality": "biomedical_literature_text",
4
+ "raw_file": "/local/home/dtabares/data/raw_embeddings/bioconceptvec/concept_cbow.json",
5
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
6
+ "original_identifier_type": "BioConceptVec concept key with Entrez/NCBI Gene ID, e.g. Gene_7157",
7
+ "mapping_strategy": "strict single-Entrez BioConceptVec gene concept mapping using enriched master table column map_entrez_ids_all",
8
+ "multi_entrez_gene_concepts": "excluded and reported",
9
+ "duplicate_source_rows": "averaged when multiple source rows mapped to the same master gene",
10
+ "ambiguous_source_rows": "excluded and reported",
11
+ "dtype_saved": "float32",
12
+ "n_total_concepts_raw": 402712,
13
+ "n_non_gene_concepts_skipped": 258128,
14
+ "n_single_entrez_gene_concepts": 126751,
15
+ "n_multi_entrez_gene_concepts_excluded": 17832,
16
+ "n_bad_gene_concepts_excluded": 1,
17
+ "n_mapped_source_rows": 17042,
18
+ "n_unmapped_source_rows": 109677,
19
+ "n_ambiguous_source_rows_excluded": 32,
20
+ "n_invalid_dim_source_rows_skipped": 0,
21
+ "n_nonfinite_source_rows_skipped": 0,
22
+ "n_master_rows": 20097,
23
+ "n_mapped_genes": 17040,
24
+ "n_genes_added_vs_strict_v1": 1,
25
+ "n_genes_lost_vs_strict_v1": 1,
26
+ "embedding_dim": 100,
27
+ "coverage_of_master_rows": 0.8478877444394686,
28
+ "n_zero_vectors": 0,
29
+ "output_npz": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_embeddings.npz",
30
+ "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_cbow_v1_1/bioconceptvec_cbow_v1_1_genes.tsv"
31
+ }
data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:631914ec43d7115189efe20cf016aa58f3740e03d782045ef5334002fe7cf111
3
+ size 6371453
data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_metadata.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "BioConceptVec-fastText-v1.1",
3
+ "modality": "biomedical_literature_text",
4
+ "raw_file": "/local/home/dtabares/data/raw_embeddings/bioconceptvec/concept_fast.json",
5
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
6
+ "original_identifier_type": "BioConceptVec concept key with Entrez/NCBI Gene ID, e.g. Gene_7157",
7
+ "mapping_strategy": "strict single-Entrez BioConceptVec gene concept mapping using enriched master table column map_entrez_ids_all",
8
+ "multi_entrez_gene_concepts": "excluded and reported",
9
+ "duplicate_source_rows": "averaged when multiple source rows mapped to the same master gene",
10
+ "ambiguous_source_rows": "excluded and reported",
11
+ "dtype_saved": "float32",
12
+ "n_total_concepts_raw": 402712,
13
+ "n_non_gene_concepts_skipped": 258128,
14
+ "n_single_entrez_gene_concepts": 126751,
15
+ "n_multi_entrez_gene_concepts_excluded": 17832,
16
+ "n_bad_gene_concepts_excluded": 1,
17
+ "n_mapped_source_rows": 17042,
18
+ "n_unmapped_source_rows": 109677,
19
+ "n_ambiguous_source_rows_excluded": 32,
20
+ "n_invalid_dim_source_rows_skipped": 0,
21
+ "n_nonfinite_source_rows_skipped": 0,
22
+ "n_master_rows": 20097,
23
+ "n_mapped_genes": 17040,
24
+ "n_genes_added_vs_strict_v1": 1,
25
+ "n_genes_lost_vs_strict_v1": 1,
26
+ "embedding_dim": 100,
27
+ "coverage_of_master_rows": 0.8478877444394686,
28
+ "n_zero_vectors": 0,
29
+ "output_npz": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_embeddings.npz",
30
+ "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_fasttext_v1_1/bioconceptvec_fasttext_v1_1_genes.tsv"
31
+ }
data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5af25c84bb0bce6ba29b8e8aaf5bb10ada95dc0cc51a3a29137cce1606e2df55
3
+ size 6466670
data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_metadata.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "BioConceptVec-GloVe-v1.1",
3
+ "modality": "biomedical_literature_text",
4
+ "raw_file": "/local/home/dtabares/data/raw_embeddings/bioconceptvec/concept_glove.json",
5
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
6
+ "original_identifier_type": "BioConceptVec concept key with Entrez/NCBI Gene ID, e.g. Gene_7157",
7
+ "mapping_strategy": "strict single-Entrez BioConceptVec gene concept mapping using enriched master table column map_entrez_ids_all",
8
+ "multi_entrez_gene_concepts": "excluded and reported",
9
+ "duplicate_source_rows": "averaged when multiple source rows mapped to the same master gene",
10
+ "ambiguous_source_rows": "excluded and reported",
11
+ "dtype_saved": "float32",
12
+ "n_total_concepts_raw": 402712,
13
+ "n_non_gene_concepts_skipped": 258128,
14
+ "n_single_entrez_gene_concepts": 126751,
15
+ "n_multi_entrez_gene_concepts_excluded": 17832,
16
+ "n_bad_gene_concepts_excluded": 1,
17
+ "n_mapped_source_rows": 17042,
18
+ "n_unmapped_source_rows": 109677,
19
+ "n_ambiguous_source_rows_excluded": 32,
20
+ "n_invalid_dim_source_rows_skipped": 0,
21
+ "n_nonfinite_source_rows_skipped": 0,
22
+ "n_master_rows": 20097,
23
+ "n_mapped_genes": 17040,
24
+ "n_genes_added_vs_strict_v1": 1,
25
+ "n_genes_lost_vs_strict_v1": 1,
26
+ "embedding_dim": 100,
27
+ "coverage_of_master_rows": 0.8478877444394686,
28
+ "n_zero_vectors": 0,
29
+ "output_npz": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_embeddings.npz",
30
+ "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_glove_v1_1/bioconceptvec_glove_v1_1_genes.tsv"
31
+ }
data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe5ed057f62fa46b31289caf8e552d754fc0732fd92bcb8740e46df5efdd9d11
3
+ size 6451563
data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_metadata.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "BioConceptVec-SkipGram-v1.1",
3
+ "modality": "biomedical_literature_text",
4
+ "raw_file": "/local/home/dtabares/data/raw_embeddings/bioconceptvec/concept_skip.json",
5
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
6
+ "original_identifier_type": "BioConceptVec concept key with Entrez/NCBI Gene ID, e.g. Gene_7157",
7
+ "mapping_strategy": "strict single-Entrez BioConceptVec gene concept mapping using enriched master table column map_entrez_ids_all",
8
+ "multi_entrez_gene_concepts": "excluded and reported",
9
+ "duplicate_source_rows": "averaged when multiple source rows mapped to the same master gene",
10
+ "ambiguous_source_rows": "excluded and reported",
11
+ "dtype_saved": "float32",
12
+ "n_total_concepts_raw": 402712,
13
+ "n_non_gene_concepts_skipped": 258128,
14
+ "n_single_entrez_gene_concepts": 126751,
15
+ "n_multi_entrez_gene_concepts_excluded": 17832,
16
+ "n_bad_gene_concepts_excluded": 1,
17
+ "n_mapped_source_rows": 17042,
18
+ "n_unmapped_source_rows": 109677,
19
+ "n_ambiguous_source_rows_excluded": 32,
20
+ "n_invalid_dim_source_rows_skipped": 0,
21
+ "n_nonfinite_source_rows_skipped": 0,
22
+ "n_master_rows": 20097,
23
+ "n_mapped_genes": 17040,
24
+ "n_genes_added_vs_strict_v1": 1,
25
+ "n_genes_lost_vs_strict_v1": 1,
26
+ "embedding_dim": 100,
27
+ "coverage_of_master_rows": 0.8478877444394686,
28
+ "n_zero_vectors": 0,
29
+ "output_npz": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_embeddings.npz",
30
+ "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/bioconceptvec_skipgram_v1_1/bioconceptvec_skipgram_v1_1_genes.tsv"
31
+ }
data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8fc73d9397b2272bfdf0f8ed277ddcfeb2cecccc356079d19e719fd6808448f
3
+ size 14768959
data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1_metadata.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "CMap-GSE106127-CGS-mean-landmark978-zscore-v1.1",
3
+ "status": "processed_qc_passed",
4
+ "modality": "perturbational transcriptomics / L1000 consensus genetic perturbation signatures",
5
+ "species": "human",
6
+ "source": "CMap/L1000 GEO GSE106127 CGS_n33839x978",
7
+ "source_gctx_file": "/local/home/dtabares/data/raw_embeddings/cmap_gse106127_cgs_v1_1/GSE106127_CGS_n33839x978.gctx",
8
+ "source_metadata_file": "/local/home/dtabares/data/raw_embeddings/cmap_gse106127_cgs_v1_1/GSE106127_CGS_meta.txt.gz",
9
+ "source_gene_info_file": "/local/home/dtabares/data/raw_embeddings/cmap_gse106127_cgs_v1_1/GSE106127_gene_info.txt.gz",
10
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
11
+ "identifier_mapping": "CMap pert_iname target symbols mapped case-insensitively to master gene_symbol.",
12
+ "preprocessing": "GCTX matrix interpreted as CGS signatures x 978 landmark response genes. Signatures were filtered to pert_type trt_sh.cgs, averaged by perturbed gene symbol across cell lines/signatures, then landmark response gene columns were z-scored across mapped perturbed genes.",
13
+ "aggregation": "mean across all CGS signatures for each perturbed gene symbol",
14
+ "embedding_shape": [
15
+ 4070,
16
+ 978
17
+ ],
18
+ "embedding_dimension": 978,
19
+ "n_raw_signatures": 33839,
20
+ "n_cgs_signatures": 33503,
21
+ "n_landmark_response_genes": 978,
22
+ "n_unique_perturbed_gene_symbols": 4312,
23
+ "n_cell_lines": 9,
24
+ "n_master_genes": 20097,
25
+ "n_mapped_master_genes": 4070,
26
+ "n_unmapped_master_genes": 16027,
27
+ "n_unmapped_source_gene_symbols": 249,
28
+ "output_files": {
29
+ "embeddings": "/local/home/dtabares/data/processed_embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/embeddings.npy",
30
+ "genes": "/local/home/dtabares/data/processed_embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/genes.tsv",
31
+ "landmark_response_gene_columns": "/local/home/dtabares/data/processed_embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/landmark_response_gene_columns.tsv",
32
+ "target_gene_signature_counts": "/local/home/dtabares/data/processed_embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/target_gene_signature_counts.tsv",
33
+ "metadata": "/local/home/dtabares/data/processed_embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/metadata.json",
34
+ "mapping_report": "/local/home/dtabares/data/processed_embeddings/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/mapping_report.txt",
35
+ "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/unmapped_master_genes.tsv",
36
+ "unmapped_source_gene_symbols": "/local/home/dtabares/reports/mapping_reports/cmap_gse106127_cgs_mean_landmark978_zscore_v1_1/unmapped_source_gene_symbols.tsv"
37
+ }
38
+ }
data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68eef8e849466aca1dfcfaa288143fdd63c3b1f597eda181a390f0ce277193fc
3
+ size 18405999
data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_metadata.json ADDED
@@ -0,0 +1,148 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "deepNF-STRING-v12-PPMI-SVD-fusion-v1.1",
3
+ "source_embedding": "deepNF-STRING-v12-PPMI-SVD-fusion-v1.1",
4
+ "core_status": "candidate_core_pending_supervisor_approval",
5
+ "modality": "STRING network multimodal sparse PPMI/SVD fusion",
6
+ "source_identifier_type": "STRING protein ID / Ensembl protein ID",
7
+ "download_source": "STRING v12 human protein.links.detailed, protein.info, and protein.aliases files",
8
+ "original_model_or_method": "deepNF: deep network fusion for protein function prediction",
9
+ "embedding_generated_by": "Daniel locally from STRING v12 channel networks",
10
+ "exact_original_reproduction": false,
11
+ "exact_original_reproduction_note": "The original deepNF data link returned 404, no archived Wayback files were found, GitHub issues report the data link unavailable, and inspected forks did not contain the original matrices. This is a stable regenerated deepNF-derived representation using sparse PPMI and SVD fusion.",
12
+ "input_data_or_sequence_source": "STRING v12 Homo sapiens evidence-channel network scores",
13
+ "string_channels": [
14
+ "coexpression",
15
+ "cooccurence",
16
+ "database",
17
+ "experimental",
18
+ "fusion",
19
+ "neighborhood",
20
+ "textmining"
21
+ ],
22
+ "min_channel_score": 150,
23
+ "per_channel_svd_dim": 128,
24
+ "final_svd_fusion_dim": 256,
25
+ "pipeline": [
26
+ "Split STRING v12 detailed network into evidence channels",
27
+ "Build one sparse weighted adjacency per evidence channel",
28
+ "Compute sparse PPMI on nonzero entries",
29
+ "L2-normalize sparse PPMI rows",
30
+ "TruncatedSVD per channel to 128 dimensions",
31
+ "Concatenate seven channel embeddings to 896 dimensions",
32
+ "Final TruncatedSVD fusion to 256 dimensions",
33
+ "Standardize final dimensions",
34
+ "Map STRING protein nodes to master genes and average duplicate STRING proteins mapping to the same gene"
35
+ ],
36
+ "final_svd": {
37
+ "n_components": 256,
38
+ "n_iter": 10,
39
+ "random_state": 42,
40
+ "explained_variance_ratio_sum": 0.48400235176086426
41
+ },
42
+ "source_files": {
43
+ "string_links_detailed": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/9606.protein.links.detailed.v12.0.txt.gz",
44
+ "string_info": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/9606.protein.info.v12.0.txt.gz",
45
+ "string_aliases": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/9606.protein.aliases.v12.0.txt.gz",
46
+ "edgelist_dir": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/deepnf_edgelists"
47
+ },
48
+ "channel_summaries": [
49
+ {
50
+ "channel": "coexpression",
51
+ "edgelist": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/deepnf_edgelists/string_v12_coexpression_min150.edgelist",
52
+ "edges_input": 895311,
53
+ "raw_adjacency_nnz_symmetric": 1790622,
54
+ "ppmi_nnz": 1789428,
55
+ "nodes_total": 19619,
56
+ "per_channel_dim": 128,
57
+ "svd_explained_variance_ratio_sum": 0.22203025221824646
58
+ },
59
+ {
60
+ "channel": "cooccurence",
61
+ "edgelist": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/deepnf_edgelists/string_v12_cooccurence_min150.edgelist",
62
+ "edges_input": 67734,
63
+ "raw_adjacency_nnz_symmetric": 135468,
64
+ "ppmi_nnz": 135468,
65
+ "nodes_total": 19619,
66
+ "per_channel_dim": 128,
67
+ "svd_explained_variance_ratio_sum": 0.15942813456058502
68
+ },
69
+ {
70
+ "channel": "database",
71
+ "edgelist": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/deepnf_edgelists/string_v12_database_min150.edgelist",
72
+ "edges_input": 295505,
73
+ "raw_adjacency_nnz_symmetric": 591010,
74
+ "ppmi_nnz": 590954,
75
+ "nodes_total": 19619,
76
+ "per_channel_dim": 128,
77
+ "svd_explained_variance_ratio_sum": 0.4126427173614502
78
+ },
79
+ {
80
+ "channel": "experimental",
81
+ "edgelist": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/deepnf_edgelists/string_v12_experimental_min150.edgelist",
82
+ "edges_input": 860467,
83
+ "raw_adjacency_nnz_symmetric": 1720934,
84
+ "ppmi_nnz": 1715606,
85
+ "nodes_total": 19619,
86
+ "per_channel_dim": 128,
87
+ "svd_explained_variance_ratio_sum": 0.22443903982639313
88
+ },
89
+ {
90
+ "channel": "fusion",
91
+ "edgelist": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/deepnf_edgelists/string_v12_fusion_min150.edgelist",
92
+ "edges_input": 7877,
93
+ "raw_adjacency_nnz_symmetric": 15754,
94
+ "ppmi_nnz": 15754,
95
+ "nodes_total": 19619,
96
+ "per_channel_dim": 128,
97
+ "svd_explained_variance_ratio_sum": 0.12593072652816772
98
+ },
99
+ {
100
+ "channel": "neighborhood",
101
+ "edgelist": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/deepnf_edgelists/string_v12_neighborhood_min150.edgelist",
102
+ "edges_input": 4599,
103
+ "raw_adjacency_nnz_symmetric": 9198,
104
+ "ppmi_nnz": 8896,
105
+ "nodes_total": 19619,
106
+ "per_channel_dim": 128,
107
+ "svd_explained_variance_ratio_sum": 0.9010187387466431
108
+ },
109
+ {
110
+ "channel": "textmining",
111
+ "edgelist": "/local/home/dtabares/data/raw_embeddings/deepnf_original_try/string_v12/deepnf_edgelists/string_v12_textmining_min150.edgelist",
112
+ "edges_input": 3532476,
113
+ "raw_adjacency_nnz_symmetric": 7064952,
114
+ "ppmi_nnz": 7044370,
115
+ "nodes_total": 19619,
116
+ "per_channel_dim": 128,
117
+ "svd_explained_variance_ratio_sum": 0.19108518958091736
118
+ }
119
+ ],
120
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
121
+ "mapping_strategy": "Map STRING protein nodes to master genes using exact STRING protein IDs, Ensembl protein IDs, STRING alias Entrez IDs, and preferred symbols. Only unique final master mappings are accepted. Multiple STRING protein nodes mapping to the same master gene are averaged.",
122
+ "counts": {
123
+ "source_string_nodes": 19619,
124
+ "source_concat_dimensions": 896,
125
+ "source_final_dimensions": 256,
126
+ "mapped_source_nodes_before_duplicate_collapse": 19101,
127
+ "unmapped_source_nodes": 152,
128
+ "ambiguous_source_nodes": 366,
129
+ "final_mapped_genes": 19071,
130
+ "final_dimensions": 256,
131
+ "master_rows": 20097,
132
+ "duplicate_final_mappings_averaged": 29
133
+ },
134
+ "coverage": {
135
+ "source_node_mapping_fraction": 0.9735970232937459,
136
+ "master_gene_coverage_fraction": 0.9489476041200179
137
+ },
138
+ "outputs": {
139
+ "npz": "/local/home/dtabares/data/processed_embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_embeddings.npz",
140
+ "genes_tsv": "/local/home/dtabares/data/processed_embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_genes.tsv",
141
+ "source_nodes_tsv": "/local/home/dtabares/data/processed_embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_source_string_nodes.tsv",
142
+ "metadata_json": "/local/home/dtabares/data/processed_embeddings/deepnf_string_v12_ppmi_svd_fusion_v1_1/deepnf_string_v12_ppmi_svd_fusion_v1_1_metadata.json",
143
+ "mapping_report": "/local/home/dtabares/reports/mapping_reports/deepnf_string_v12_ppmi_svd_fusion_v1_1_mapping_report.txt",
144
+ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/deepnf_string_v12_ppmi_svd_fusion_v1_1_unmapped_source_identifiers.tsv",
145
+ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/deepnf_string_v12_ppmi_svd_fusion_v1_1_ambiguous_source_identifiers.tsv",
146
+ "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/deepnf_string_v12_ppmi_svd_fusion_v1_1_duplicate_final_mappings.tsv"
147
+ }
148
+ }
data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ba554fd49106c5b9fef61f9612cf68e586a63aee6cdbb430d23d3c85c02b730
3
+ size 66380222
data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/deepsea_tss_1kb_epigenomic_probs_v1_1/deepsea_tss_1kb_epigenomic_probs_v1_1_metadata.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "DeepSEA-TSS-1kb-epigenomic-probabilities-zscore-v1.1",
3
+ "status": "processed_qc_passed",
4
+ "modality": "regulatory DNA / promoter-TSS sequence model",
5
+ "model": "DeepSEA/predict from Kipoi",
6
+ "model_source": "Kipoi model zoo; DeepSEA converted PyTorch model",
7
+ "raw_output": "919 DeepSEA epigenomic probabilities per TSS-centered 1 kb sequence",
8
+ "final_embedding": "dimension-wise z-scored DeepSEA probabilities across mapped genes",
9
+ "input_sequence_type": "genomic DNA TSS-centered 1 kb window",
10
+ "input_sequence_source": "/local/home/dtabares/data/raw_embeddings/deepsea_tss_1kb_v1_1/reference_genome/GRCh38.primary_assembly.genome.fa",
11
+ "genome_build": "GRCh38",
12
+ "annotation_source": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
13
+ "tss_rule": "plus strand: gene_start; minus strand: gene_end; coordinates from master_gene_table_v1_1_enriched.csv",
14
+ "window_rule": "BED interval [TSS0 - 500, TSS0 + 500), where TSS0 is 0-based TSS",
15
+ "strand_handling": "use_strand=False; sequences are genomic plus-reference orientation, not gene-oriented reverse-complemented",
16
+ "dimension": 919,
17
+ "n_master_rows_input": 20097,
18
+ "n_valid_tss_windows": 20097,
19
+ "n_skipped_tss_windows": 0,
20
+ "output_files": {
21
+ "embeddings": "embeddings.npy",
22
+ "raw_probabilities": "raw_deepsea_probabilities.npy",
23
+ "genes": "genes.tsv",
24
+ "bed": "deepsea_tss_1kb_windows.bed",
25
+ "predictions_tsv": "deepsea_tss_1kb_predictions.tsv",
26
+ "feature_columns": "deepsea_feature_columns.tsv",
27
+ "skipped_windows": "/local/home/dtabares/reports/mapping_reports/deepsea_tss_1kb_epigenomic_probs_v1_1/skipped_tss_windows.tsv"
28
+ },
29
+ "created_at": "2026-07-09T10:41:32"
30
+ }
data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:991d7bfc533eba950f0b576406c65ea617d7a0b629d9454d8fc0622916fbd708
3
+ size 17729467
data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/depmap_crispr_gene_effect_pca256_v1_1/depmap_crispr_gene_effect_pca256_v1_1_metadata.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "DepMap-CRISPRGeneEffect-PCA256-v1.1",
3
+ "status": "processed_qc_passed",
4
+ "modality": "functional genomics / CRISPR dependency",
5
+ "species": "human",
6
+ "source": "DepMap CRISPRGeneEffect.csv",
7
+ "source_file": "/local/home/dtabares/data/raw_embeddings/depmap_crispr_gene_effect_v1_1/CRISPRGeneEffect.csv",
8
+ "model_metadata_file": "/local/home/dtabares/data/raw_embeddings/depmap_crispr_gene_effect_v1_1/Model.csv",
9
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
10
+ "identifier_mapping": "DepMap columns parsed as SYMBOL (Entrez); mapped to master entrez_id.",
11
+ "preprocessing": "Gene-effect matrix transposed to genes x models; missing values filled by source gene mean; model/cell-line features standardized across mapped genes; PCA with randomized SVD.",
12
+ "pca_n_components": 256,
13
+ "embedding_shape": [
14
+ 18562,
15
+ 256
16
+ ],
17
+ "embedding_dimension": 256,
18
+ "n_depmap_models": 1208,
19
+ "n_source_gene_columns_parsed": 18531,
20
+ "n_source_gene_columns_unparsed": 0,
21
+ "n_unique_source_entrez_after_collapse": 18531,
22
+ "duplicate_source_entrez_ids_collapsed_by_mean": 0,
23
+ "n_master_genes": 20097,
24
+ "n_mapped_master_genes": 18562,
25
+ "n_unmapped_master_genes": 1535,
26
+ "pca_total_explained_variance_ratio": 0.929381850584523,
27
+ "output_files": {
28
+ "embeddings": "/local/home/dtabares/data/processed_embeddings/depmap_crispr_gene_effect_pca256_v1_1/embeddings.npy",
29
+ "genes": "/local/home/dtabares/data/processed_embeddings/depmap_crispr_gene_effect_pca256_v1_1/genes.tsv",
30
+ "model_columns": "/local/home/dtabares/data/processed_embeddings/depmap_crispr_gene_effect_pca256_v1_1/model_columns.tsv",
31
+ "pca_explained_variance": "/local/home/dtabares/data/processed_embeddings/depmap_crispr_gene_effect_pca256_v1_1/pca_explained_variance.tsv",
32
+ "metadata": "/local/home/dtabares/data/processed_embeddings/depmap_crispr_gene_effect_pca256_v1_1/metadata.json",
33
+ "mapping_report": "/local/home/dtabares/data/processed_embeddings/depmap_crispr_gene_effect_pca256_v1_1/mapping_report.txt",
34
+ "unmapped_master_genes": "/local/home/dtabares/reports/mapping_reports/depmap_crispr_gene_effect_pca256_v1_1/unmapped_master_genes.tsv"
35
+ }
36
+ }
data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34bd75200f6e165ff84ed2de022c0f32989470a373c267385d5505757a16e045
3
+ size 57063502
data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/dnabert2_117m_cds_v1_1/dnabert2_117m_cds_v1_1_metadata.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "dnabert2_117m_cds_v1_1",
3
+ "status": "processed_qc_passed",
4
+ "created_at": "2026-07-08T12:55:46",
5
+ "model": "DNABERT-2 117M",
6
+ "model_source": "zhihan1996/DNABERT-2-117M",
7
+ "model_directory": "/local/home/dtabares/data/raw_embeddings/dnabert2_117m_cds_v1_1/hf_model",
8
+ "model_weight_file": "/local/home/dtabares/data/raw_embeddings/dnabert2_117m_cds_v1_1/hf_model/pytorch_model.bin",
9
+ "model_weight_sha256": "7ff39ec77a484dd01070a41bfd6e95cdd7247bec80fe357ab43a4be33687aeba",
10
+ "embedding_dimension": 768,
11
+ "matrix_shape": [
12
+ 20085,
13
+ 768
14
+ ],
15
+ "input_sequence_type": "canonical protein-coding CDS DNA sequence",
16
+ "input_sequence_file": "/local/home/dtabares/data/raw_embeddings/dnabert2_117m_cds_v1_1/source_sequences/dnabert2_canonical_cds_sequences.tsv",
17
+ "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.",
18
+ "chunk_nt": 1600,
19
+ "max_tokens": 512,
20
+ "chunked_genes": 3505,
21
+ "attention_workaround": {
22
+ "reason": "Official DNABERT-2 custom Triton FlashAttention path hangs on this workstation. Setting attention_probs_dropout_prob=0.1 routes to the PyTorch attention implementation while model.eval() disables stochastic dropout.",
23
+ "original_attention_probs_dropout_prob": 0.0,
24
+ "forced_attention_probs_dropout_prob": 0.1
25
+ },
26
+ "official_repo": "/local/home/dtabares/data/raw_embeddings/dnabert2_117m_cds_v1_1/official_repo",
27
+ "official_repo_commit": null,
28
+ "master_gene_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
29
+ "software": {
30
+ "torch": "2.0.1+cu118",
31
+ "transformers": "4.29.2",
32
+ "numpy": "1.26.4"
33
+ },
34
+ "runtime_seconds": 1222.18
35
+ }
data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:173de92296209c963ac47269e5cf3125f3da9d2d2c768a2c676ddfa33d38c703
3
+ size 390033274
data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/enformer_tss_393kb_human_central3_v1_1/enformer_tss_393kb_human_central3_v1_1_metadata.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "Enformer-TSS-393kb-human-central3-log1p-zscore-v1.1",
3
+ "status": "processed_qc_passed",
4
+ "modality": "regulatory DNA / long-range TSS sequence model",
5
+ "model": "Enformer",
6
+ "model_source": "https://tfhub.dev/deepmind/enformer/1",
7
+ "kept_output_head": "human only",
8
+ "discarded_output_head": "mouse",
9
+ "human_output_shape_per_sequence": [
10
+ 896,
11
+ 5313
12
+ ],
13
+ "pooled_representation": "mean of central 3 human output bins around TSS",
14
+ "raw_output_file": "raw_enformer_human_central3_mean_tracks.npy",
15
+ "transformation": "log1p raw human central3 mean tracks, then dimension-wise z-score across genes",
16
+ "final_embedding_file": "embeddings.npy",
17
+ "input_sequence_type": "GRCh38 genomic DNA, 393216 bp centered on gene TSS",
18
+ "input_sequence_source": "/local/home/dtabares/data/raw_embeddings/deepsea_tss_1kb_v1_1/reference_genome/GRCh38.primary_assembly.genome.fa",
19
+ "annotation_source": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
20
+ "genome_build": "GRCh38",
21
+ "sequence_length_bp": 393216,
22
+ "target_length_bins": 896,
23
+ "central_bins_used_0based": [
24
+ 447,
25
+ 448,
26
+ 449
27
+ ],
28
+ "tss_rule": "plus strand: gene_start; minus strand: gene_end; coordinates from master_gene_table_v1_1_enriched.csv",
29
+ "strand_handling": "genomic reference orientation; no gene-strand reverse-complementing",
30
+ "boundary_handling": "N-padding if the 393216 bp window extends beyond chromosome boundaries",
31
+ "n_master_rows_input": 20097,
32
+ "n_valid_tss_windows": 20097,
33
+ "n_skipped_tss_windows": 0,
34
+ "dimension": 5313,
35
+ "created_at": "2026-07-10T06:03:35"
36
+ }
data/embeddings/esm2_v1_1/esm2_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:526110c9441d080b536c7e6a70bc036692ac1e84855fca9abffead9bfdbee790
3
+ size 88942773
data/embeddings/esm2_v1_1/esm2_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/esm2_v1_1/esm2_v1_1_metadata.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "ESM2-v1.1",
3
+ "source_embedding": "Zhong_all_genes_ESM2",
4
+ "modality": "protein_sequence",
5
+ "algorithm": "ESM2 Transformer",
6
+ "source_identifier_type": "Entrez",
7
+ "source_files": {
8
+ "embedding_matrix": "/local/home/dtabares/data/raw_embeddings/zhong_gene_embedding_benchmark/full_embeddings/all_genes/ESM2/ESM2_emb.csv",
9
+ "genelist": "/local/home/dtabares/data/raw_embeddings/zhong_gene_embedding_benchmark/full_embeddings/all_genes/ESM2/ESM2_genelist.txt"
10
+ },
11
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
12
+ "mapping_strategy": "Exact Entrez ID mapping to master_gene_table_v1_1_enriched.csv. Only Entrez IDs mapping to exactly one master gene are accepted. Duplicate source rows mapping to the same final master gene are averaged.",
13
+ "counts": {
14
+ "source_rows": 18903,
15
+ "source_dimensions": 1280,
16
+ "mapped_source_rows_before_duplicate_collapse": 18708,
17
+ "unmapped_source_rows": 35,
18
+ "ambiguous_source_rows": 160,
19
+ "final_mapped_genes": 18690,
20
+ "final_dimensions": 1280,
21
+ "master_rows": 20097,
22
+ "duplicate_final_mappings_averaged": 16
23
+ },
24
+ "coverage": {
25
+ "source_row_mapping_fraction": 0.9896841771147437,
26
+ "master_gene_coverage_fraction": 0.9299895506792059
27
+ },
28
+ "outputs": {
29
+ "npz": "/local/home/dtabares/data/processed_embeddings/esm2_v1_1/esm2_v1_1_embeddings.npz",
30
+ "genes_tsv": "/local/home/dtabares/data/processed_embeddings/esm2_v1_1/esm2_v1_1_genes.tsv",
31
+ "metadata_json": "/local/home/dtabares/data/processed_embeddings/esm2_v1_1/esm2_v1_1_metadata.json",
32
+ "mapping_report": "/local/home/dtabares/reports/mapping_reports/esm2_v1_1_mapping_report.txt",
33
+ "unmapped_tsv": "/local/home/dtabares/reports/mapping_reports/esm2_v1_1_unmapped_source_identifiers.tsv",
34
+ "ambiguous_tsv": "/local/home/dtabares/reports/mapping_reports/esm2_v1_1_ambiguous_source_identifiers.tsv",
35
+ "duplicates_tsv": "/local/home/dtabares/reports/mapping_reports/esm2_v1_1_duplicate_final_mappings.tsv"
36
+ }
37
+ }
data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ee9ae0e584204bab577e64a72d089877dd5da62faf8c47ec4ea5b6368089fba
3
+ size 15433953
data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/frogs_archs4_256_v1_1/frogs_archs4_256_v1_1_metadata.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "frogs_archs4_256_v1_1",
3
+ "status": "processed_qc_passed",
4
+ "created_at": "2026-07-08T13:32:49",
5
+ "model": "FRoGS ARCHS4 gene embedding",
6
+ "model_source": "chenhcs/FRoGS data/gene_vec_archs4_256.csv",
7
+ "official_repo": "/local/home/dtabares/data/raw_embeddings/frogs_archs4_256_v1_1/official_repo",
8
+ "official_repo_commit": "7f2f68340b961f03c74f3e3a6c7ba6a6eb42c8c0",
9
+ "source_csv": "/local/home/dtabares/data/raw_embeddings/frogs_archs4_256_v1_1/official_repo/data/gene_vec_archs4_256.csv",
10
+ "source_csv_sha256": "dce21393bc6ad8091fc416eaf00b5df8f0226fd52a25a526799273799903c307",
11
+ "original_identifier_type": "Entrez Gene ID",
12
+ "mapping_policy": "FRoGS source Entrez IDs mapped to master Ensembl gene IDs. Entrez IDs mapping to multiple master Ensembl genes were excluded. Duplicate source Entrez IDs were averaged before mapping. Multiple source Entrez IDs mapping to the same final Ensembl gene were averaged after mapping.",
13
+ "master_gene_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
14
+ "entrez_columns_used": [
15
+ "entrez_id",
16
+ "map_entrez_ids_all",
17
+ "entrez_id_hgnc",
18
+ "entrez_id_biomart",
19
+ "hgnc_entrez_id"
20
+ ],
21
+ "embedding_dimension": 256,
22
+ "matrix_shape": [
23
+ 16265,
24
+ 256
25
+ ],
26
+ "raw_shape": [
27
+ 22970,
28
+ 257
29
+ ],
30
+ "duplicate_entrez_rows_before_averaging": 0,
31
+ "duplicated_ensembl_rows_before_final_collapse": 6,
32
+ "ambiguous_entrez_ids_excluded_from_master_mapping": 272,
33
+ "runtime_seconds": 3.64
34
+ }
data/embeddings/gene2vec_v1_1/gene2vec_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b4d577d5e0fa83e657e6e89d75b92b7c2e01060188214c548b2758a6e020dbf
3
+ size 13415164
data/embeddings/gene2vec_v1_1/gene2vec_v1_1_genes.tsv ADDED
The diff for this file is too large to render. See raw diff
 
data/embeddings/gene2vec_v1_1/gene2vec_v1_1_metadata.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embedding_name": "Gene2Vec-v1.1",
3
+ "modality": "coexpression_expression",
4
+ "raw_file": "/local/home/dtabares/data/raw_embeddings/gene2vec/gene2vec_dim_200_iter_9.txt",
5
+ "master_table": "/local/home/dtabares/metadata/master_gene_table_v1_1_enriched.csv",
6
+ "original_identifier_type": "gene_symbol",
7
+ "mapping_strategy": "priority symbol mapping using enriched master table: current gene symbols, HGNC previous symbols, HGNC alias symbols",
8
+ "duplicate_source_rows": "averaged when multiple source rows mapped to the same master gene",
9
+ "ambiguous_source_rows": "excluded and reported",
10
+ "dtype_saved": "float32",
11
+ "n_source_rows_raw": 24447,
12
+ "n_bad_rows_skipped": 0,
13
+ "n_mapped_source_rows": 17965,
14
+ "n_unmapped_source_rows": 6450,
15
+ "n_ambiguous_source_rows_excluded": 32,
16
+ "n_nonfinite_source_rows_skipped": 0,
17
+ "mapping_method_counts": {
18
+ "direct_current_gene_symbol": 16796,
19
+ "hgnc_alias_symbol": 39,
20
+ "hgnc_previous_symbol": 1130
21
+ },
22
+ "n_master_rows": 20097,
23
+ "n_mapped_genes": 17916,
24
+ "n_genes_added_vs_gene2vec_strict_v1": 1126,
25
+ "embedding_dim": 200,
26
+ "coverage_of_master_rows": 0.8914763397522019,
27
+ "n_zero_vectors": 0,
28
+ "output_npz": "/local/home/dtabares/data/processed_embeddings/gene2vec_v1_1/gene2vec_v1_1_embeddings.npz",
29
+ "output_genes_tsv": "/local/home/dtabares/data/processed_embeddings/gene2vec_v1_1/gene2vec_v1_1_genes.tsv"
30
+ }
data/embeddings/genecompass_base_gene_token_v1_1/genecompass_base_gene_token_v1_1_embeddings.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45edd21c2abc6024257dd70e47ec40658cb47e4a4ec9ca5e3be514e57acf164f
3
+ size 47272177