DisProt / dataset_summary.json
anindya64's picture
Add normalized Parquet train/test DisProt protein table
5e142fe verified
{
"source": "LiteFold/DisProt",
"entry_rows": 3199,
"region_rows": 13396,
"splits": {
"train": 2875,
"test": 324
},
"split_strategy": "deterministic sha256(disprot_id) % 10; bucket 0 is test, buckets 1-9 are train",
"dataset_label_counts": {
"NDDs-related proteins": 324,
"Viral proteins": 230,
"RNA-binding proteins": 212,
"Cancer-related proteins": 181,
"Condensates-related proteins": 181,
"Autophagy-related proteins": 109,
"Neglected tropical diseases proteins": 105,
"Extracellular matrix proteins": 100,
"Bacterial virulence-related proteins": 76,
"Unicellular toxins and antitoxins": 51,
"Age-related disorders proteins": 42
},
"top_region_terms": {
"disorder": 7138,
"protein binding": 1558,
"disorder to order": 834,
"flexible linker": 426,
"molecular adaptor activity": 262,
"phosphorylation display site": 240,
"molecular function regulator": 188,
"lipid binding": 134,
"molecular condensate scaffold activity": 106,
"order": 104,
"pre-molten globule": 94,
"RNA binding": 86,
"order to disorder": 79,
"DNA binding": 78,
"nucleic acid binding": 77,
"amyloid fibril formation": 72,
"molecular function inhibitor activity": 72,
"flexible N-terminal tail": 72,
"molecular function activator activity": 63,
"flexible C-terminal tail": 58
},
"region_namespace_counts": {
"Structural state": 7394,
"Molecular function": 3405,
"Disorder function": 1028,
"Structural transition": 965,
"Biological process": 542,
"Cellular component": 62
},
"columns": [
"disprot_id",
"accession",
"uniparc",
"name",
"organism",
"ncbi_taxon_id",
"length",
"disorder_content",
"dataset_labels",
"taxonomy",
"released",
"date",
"creator",
"uniref100",
"uniref90",
"uniref50",
"sequence",
"region_count",
"region_ids",
"region_starts",
"region_ends",
"region_lengths",
"region_terms",
"region_term_ids",
"region_namespaces",
"evidence_codes",
"reference_ids",
"reference_sources",
"curator_names",
"cross_refs",
"feature_databases",
"feature_ids",
"feature_names",
"feature_count",
"gene_names",
"consensus_starts",
"consensus_ends",
"consensus_types",
"split_bucket"
],
"metadata_tables": [
"metadata/regions.parquet"
]
}