A multiple sequence alignment algorithm for homologous proteins using secondary structure information and optionally keying alignments to functionally important sites.

Abstract

The programs described herein function as part of a suite of programs designed for pairwise alignment, multiple alignment, generation of randomized sequences, production of alignment scores and a sorting routine for analysis of the alignments produced. The sequence alignment programs penalize gaps (absences of residues) within regions of protein secondary structure and have the added option of 'fingerprinting' structurally or functionally important protein-residues. The multiple alignment program is based upon the sequence alignment method of Needleman and Wunsch and the multiple alignment extension of Barton and Sternberg. Our application includes the feature of optionally weighting active site, monomer--monomer, ligand contact or other important template residues to bias the alignment toward matching these residues. A sum-score for the alignments is introduced, which is independent of gap penalties. This score more adequately reflects the character of the alignments for a given scoring matrix than the gap-penalty-dependent total score described previously in the literature. In addition, individual amino acid similarity scores at each residue position in the alignments are printed with the alignment output to enable immediate quantitative assessment of homology at key sections of the aligned chains.