Pfam is a collection of multiple sequence alignments and hidden Markov models (i.e., profiles of the multiple alignments) of proteins. These multiple alignments define protein domains or families.

Pfam can be queried using keywords, DNA and protein sequences.