PfamScan (EBI)

PfamScan (EBI) identifies Pfam protein families and predicts active site residues in query protein sequences by scanning FASTA sequences against the Pfam Hidden Markov Models (HMMs).

Key Features:

• Pfam HMM scanning: Searches input FASTA sequences against the Pfam Hidden Markov Models (HMMs) to identify protein family matches.
• Active site prediction: Predicts active site residues in enzymatic Pfam families by transferring experimentally determined residue annotations within the same Pfam family using a strict rule set, expanding Pfam active site annotations approximately 200-fold to ~606,110 residues, of which ~94% are novel relative to UniProtKB.
• EMBL-EBI integration: Operates within the EMBL-EBI Job Dispatcher framework and integrates Pfam annotations with resources such as UniProt, InterPro, ENA, and Ensembl Genomes.
• Database updates: Refreshes active site predictions with each new Pfam release to incorporate revised family annotations.

Scientific Applications:

• Enzyme mechanism annotation: Maps putative catalytic residues to support interpretation of enzyme mechanisms.
• Functional annotation: Assigns protein family membership to query sequences to inform functional annotation and comparative analyses.
• Drug target identification: Identifies candidate catalytic or active residues to inform drug discovery and target validation.

Methodology:

Scans FASTA sequences against Pfam HMMs and predicts active site residues by transferring experimentally derived annotations across sequences within the same Pfam family using a strict rule set, with an estimated false positive rate of ~3%.