WOOF: A word-oriented objective function for validation of sequence alignments
Robert G. Beiko, Cheong Xin Chan and Mark A. Ragan
WOOF is a method designed to rigourously apply the principle of visual alignment validation. WOOF does not actually make changes to a multiple sequence alignment, but is intead intended to choose the best alignment of a set of proteins from a set of such alignments that have been generated using different algorithms and parameter settings. The winning alignment is the one that best aligns a set of conserved patterns extracted using a pattern-finding approach such as TEIRESIAS. Patterns are weighted based on their statistical significance and position relative to other patterns.
WOOF is currently implemented as a series of Perl scripts; there is room for improvement in efficiency, but the current limiting step is actually the pattern extraction step that is done by external software.
WOOF release (.tar.gz)
This archive contains the latest set of Perl scripts, documentation and a sample data set.
The citation for WOOF is:
Beiko, R.G., Chan, C.X., and Ragan, M.A. (2005) A word-oriented objective function for alignment validation. Bioinformatics 21: 2230-2239.
You can E-mail me your questions at firstname.lastname@example.org
© 2006 Robert G. Beiko