COBALT: constraint-based alignment tool for multiple protein sequences

Bioinformatics. 2007 May 1;23(9):1073-9. doi: 10.1093/bioinformatics/btm076. Epub 2007 Mar 1.

Abstract

Motivation: A tool that simultaneously aligns multiple protein sequences, automatically utilizes information about protein domains, and has a good compromise between speed and accuracy will have practical advantages over current tools.

Results: We describe COBALT, a constraint based alignment tool that implements a general framework for multiple alignment of protein sequences. COBALT finds a collection of pairwise constraints derived from database searches, sequence similarity and user input, combines these pairwise constraints, and then incorporates them into a progressive multiple alignment. We show that using constraints derived from the conserved domain database (CDD) and PROSITE protein-motif database improves COBALT's alignment quality. We also show that COBALT has reasonable runtime performance and alignment accuracy comparable to or exceeding that of other tools for a broad range of problems.

Availability: COBALT is included in the NCBI C++ toolkit. A Linux executable for COBALT, and CDD and PROSITE data used is available at: ftp://ftp.ncbi.nlm.nih.gov/pub/agarwala/cobalt

Publication types

  • Evaluation Study
  • Research Support, N.I.H., Intramural

MeSH terms

  • Algorithms*
  • Amino Acid Sequence
  • Molecular Sequence Data
  • Proteins / chemistry*
  • Sequence Alignment / methods*
  • Sequence Analysis, Protein / methods*
  • Sequence Homology, Amino Acid
  • Software*

Substances

  • Proteins