PASS2: A Database of Structure-Based Sequence Alignments of Protein Structural Domain Superfamilies

Sequence alignments guided by structural features are particularly suited for distant relationships and they permit a better sampling of the protein sequence space. Reliable sequence alignments could be useful in evolutionary biology and in defining structurefunction relationships for protein superfamilies. PASS2 database presents structure-based alignments of protein domains related at the superfamily level. The number of new superfamilies was increased by 47% compared to the previous PASS2 2003 version, which shows the crucial need for updating the PASS2 database. In the current release of PASS2, the single-member, two-member and multimember superfamilies go through different pipelines in the structural alignment protocol. We have also introduced the alignment assessment methods such as, mean-RMSD and percentage of secondary structural equivalences which should be effective in producing better alignment with minimal intervention. New features such as sequence clustering, length-conserved blocks, secondary structural equivalence and outlier-recognition are also provided. We suggest that the structurally deviant superfamily members could be removed as outliers such that extreme distant relationships will not obscure the alignment. We thus report an nearly-automated, updated version of the superfamily alignment database, consisting of 1776 superfamilies and 9536 protein domains, that is in direct correspondence with SCOP (1.73) database. The database of alignments can be accessed at http://caps.ncbs.res.in/pass2/pass2.html