Pfam contains multiple sequence alignments and profile-HMMs (hidden Markov models) of protein families and domains. Pfam consists of two parts, Pfam-A and Pfam-B. The families in Pfam-A are based on curated multiple sequence alignments, while the families in Pfam-B are derived automatically. Pfam-A contains the original seed alignment, HMM based on that, an automatic full alignment of all sequences detected by the HMM in SwissProt and TrEMBL, and a short description of the protein family.

You can also search matches to Pfam-A with your sequence using HMMER or Wise2 in Hippu. Pfam home