Multiple model approach--dealing with alignment ambiguities in protein modeling

Pac Symp Biocomput. 1997:328-39.

Abstract

Sequence alignments for distantly homologous proteins are often ambiguous, which creates a weak link in structure prediction by homology. We address this problem by using several plausible alignments in a modeling procedure, obtaining many models of the target. All are subsequently evaluated by a threading algorithm. It is shown that this approach can identify best alignments and produce reasonable models, whose quality is now limited only by the extent of the structural similarity between the known and predicted protein. Using a similar approach structure prediction for the oxidized dimer of S100A1 protein, for which the structure is not known, is presented.

Publication types

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

MeSH terms

  • Amino Acid Sequence
  • Calbindins
  • Computer Simulation*
  • Dimerization
  • Models, Molecular*
  • Molecular Sequence Data
  • Parvalbumins / chemistry
  • Protein Conformation*
  • Proteins / chemistry*
  • S100 Calcium Binding Protein G / chemistry
  • S100 Proteins / chemistry
  • Sequence Alignment
  • Sequence Homology, Amino Acid

Substances

  • Calbindins
  • Parvalbumins
  • Proteins
  • S100 Calcium Binding Protein G
  • S100 Proteins