Pfam

Pfam ( Protein Families ) is a freely accessible database for bioinformatics purposes. It is on the one hand to a machine categorization of protein domains, which includes all known proteins. It is based on pattern recognition by means of machine learning of the amino acid sequence. The determined pattern can be found again in the other new proteins, which gives an indication of the composition of proteins of domains and thus also to its function, and for enzymes the enzymic activity. For this prediction provides a web service Pfam ready.

Pfam is in two parts, A and Pfam Pfam B. In Pfam -A well-characterized domains are summarized, while domains with unknown function in Pfam -B are located. When the automatic clustering and pattern recognition method is Hidden Markov Models.

Pfam in 1997 by the bioinformaticians Erik Sonnhammer ( Karolinska Institutet in Stockholm ), Sean Eddy (Washington University in St. Louis), and Richard Durbin (Sanger Centre, Cambridge ) was established. To some functions were expanded in early 2006 came out the update 18. In March 2013 Pfam 27.0 was released.

645896
de