interpro_binding_site

Protein-level multi-label dataset sourced from InterPro entry metadata and UniProtKB protein-to-InterPro cross-references.

Intended use

Protein binding site annotation prediction from sequence. This evaluates recovery of curated InterPro classifications, not an experimental assay result.

Source and labels

• InterPro metadata: https://ftp.ebi.ac.uk/pub/databases/interpro/current_release/entry.list
• UniProt REST API: https://rest.uniprot.org/uniprotkb/stream
• Card generated (UTC): 2026-10-07
• Organism Taxonomy ID: 9606 (all = all organisms).
• Review status: reviewed (reviewed = Swiss-Prot, unreviewed = TrEMBL, all = both).
• UniProt query: (organism_id:9606) AND (reviewed:true)
• UniProt TSV fields: accession,sequence,xref_interpro,organism_id,reviewed
• InterPro entry type: Binding_site
• Consider entries observed in at least 5 proteins.
• After splitting, retain entries with at least 20 positives in training, 3 in validation, and 3 in test. Because validation and test support select the entries, scores are conditional on that label selection.
• targets is a multi-hot vector ordered by InterPro accession. Accession, display name, and type are recorded in label_vocabulary.json.
• Proteins without a retained entry are dropped, both before splitting and again after split-support filtering, so no row has all-zero targets.
• id is the UniProtKB accession of the first entry carrying each unique sequence.

Splits

Whole MMseqs2 clusters are assigned to splits targeting {'train': 0.8, 'validation': 0.1, 'test': 0.1}. Seed: 1957723. Clustering settings:

• MMseqs workflow: easy-cluster
• Minimum sequence identity (--min-seq-id): 0.3
• Minimum coverage (-c): 0.8
• Coverage mode (--cov-mode): 1
• Threads: 1
• Other MMseqs2 options use MMseqs2 defaults.

When create_split_subsets is enabled, pooled random, stratified (by each protein's globally rarest positive entry), and hold-cluster-out subsets are also included. Hold-cluster-out subsets require enough MMseqs clusters to populate all three roles. The optional maximum sequence length is None and is applied before vocabulary construction and MMseqs2 clustering.

Split sizes

• test: 58 rows
• test_cluster: 58 rows
• test_random: 58 rows
• test_stratified: 58 rows
• train: 436 rows
• train_cluster: 436 rows
• train_random: 436 rows
• train_stratified: 436 rows
• validation: 61 rows
• validation_cluster: 61 rows
• validation_random: 61 rows
• validation_stratified: 61 rows

Vocabulary size: 2.

Dataset statistics

stats.json at the dataset root contains row counts by split, columns, and SeqKit sequence-length metrics.