Pfam

Pfam is a comprehensive database of protein families, each represented by multiple sequence alignments and profile hidden Markov models (HMMs). It serves as a resource for identifying conserved domains within protein sequences and for inferring functional and evolutionary relationships among proteins.

Pfam is organized into Pfam-A, a curated set of high-quality seed alignments and their corresponding HMMs, and Pfam-B, automatically generated families that extend coverage. Each family receives a unique accession, typically starting with PF (for example PF00001), a descriptive name, and a seed alignment, a full alignment, and an HMM. Related families can be grouped into clans when there is evidence of a common evolutionary origin.

Proteins are scanned against the Pfam HMMs using tools such as HMMER to annotate domain content and predict architecture. Pfam entries commonly link to related resources and databases, including InterPro, UniProt, and the Protein Data Bank (PDB), facilitating cross-references and functional interpretation. The database is widely used in genome annotation, protein function prediction, and comparative genomics.

Access and updates: Pfam is freely accessible online, with downloadable HMMs and alignments for programmatic use. The resource is updated periodically to incorporate new sequences, revised alignments, and additional families. Pfam remains a standard reference in protein domain annotation and in studies of protein evolution and function.