Post-processing long pairwise alignments

Bioinformatics. 1999 Dec;15(12):1012-9. doi: 10.1093/bioinformatics/15.12.1012.


Motivation: The local alignment problem for two sequences requires determining similar regions, one from each sequence, and aligning those regions. For alignments computed by dynamic programming, current approaches for selecting similar regions may have potential flaws. For instance, the criterion of Smith and Waterman can lead to inclusion of an arbitrarily poor internal segment. Other approaches can generate an alignment scoring less than some of its internal segments.

Results: We develop an algorithm that decomposes a long alignment into sub-alignments that avoid these potential imperfections. Our algorithm runs in time proportional to the original alignment's length. Practical applications to alignments of genomic DNA sequences are described.

Publication types

  • Research Support, U.S. Gov't, Non-P.H.S.
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Algorithms*
  • Animals
  • Humans
  • Mice
  • Mutation
  • Sequence Alignment*
  • Sequence Analysis, Protein / methods*
  • Software*