ABSTRACT
Phil Bradley, Lenore Cowen, Matthew Menke, Jonathan King and Bonnie Berger (2001)
BetaWrap: A program for efficiently predicting ß-helices from amino acid sequence [by strand alighment using pairwise-residue correlations]
The amino acid sequence rules that specify ß-sheet structure in proteins remain obscure (Koehl and Levitt). A subclass of ß-sheet proteins, parallel ß-helices, represent a processive folding of the chain into an elongated topologically simpler fold than globular ß-sheets. We have developed a computational approach using ß-strand interactions to predict the right-handed ß-helix super-secondary structural motif from primary amino acid sequences. The program BetaWrap recognizes each of the seven known SCOP ß-helix families (Scop-Murzin), when trained on the known ß-helices from outside the family. BetaWrap identifies 2448 sequences among 595,890 screened from the NCBI nonredundant protein database as likely ß-helices. Many of these are bacterial proteins of unknown structure that lay a role in human infectious disease; these proteins include virulence factors, adhesins, and toxins in pahogenesis, as well as surface proteins from Chlamydia and the intestinal bacterium Helicobacter pylori. The computational method introducd here is termed a 3D dynamic profile method because it generates inter-strand pairwise correlations from a processive sequence wrap. Such methods may be applicable to recognizing other beta structures for which strand topology and profiles of residue accessibility are well conserved.
