Brutlag, D. L. and Klingler, T. M. (1993). Automatic Discovery of Protein Motifs. In Proceedings of the Ninth International Biotechnology Symposium, Crystal City, Virginia: ACS Books.
Department of
Biochemistry &
Section on Medical
Informatics,
Stanford University
School of Medicine, Stanford,
California 94305-5307.
We have developed a novel representation of protein motifs that
permits the rapid discovery of structural features in sets of protein
sequences with a common structure or function. Many popular methods
for representing protein motifs (consensus sequences, weight matrices
and profiles) all emphasize conservation of amino acids at specific
sites in the sequence. Our method looks for correlations between
amino acid residues at distinct sites. Correlations between the
residues represent side chain-side chain interactions and give
insight into the structural properties of the motifs. Structural
correlations can be used in database search to discover other
proteins bearing similar relationships. This database search is
significantly more sensitive than methods depending only upon
conserved residues in protein motifs.
To
get Acrobat: 
[Back to Doug] [Address] [Academics] [Honors] [Publications] [Presentations] [Public Service]