PASS2 version 4: An update to the database of structure-based sequence alignments of structural domain superfamilies

Accurate structure-based sequence alignments of distantly related proteins are crucial in gaining insight about protein domains that belong to a superfamily. The PASS2 database provides alignments of proteins related at the superfamily level and are characterized by low sequence identity. We thus report an automated, updated version of the superfamily alignment database known as PASS2.4, consisting of 1961 superfamilies and 10,569 protein domains, which is in direct correspondence with SCOP (1.75) database. Database organization, improved methods for efficient structure-based sequence alignments and the analysis of extreme distantly related proteins within superfamilies formed the focus of this update. Alignment of family-specific functional residues can be realized using such alignments and is shown using one superfamily as an example. The database of alignments and other related features can be accessed at http://caps.ncbs.res.in/pass2/.