Format

Send to

Choose Destination
See comment in PubMed Commons below
Nat Struct Biol. 1999 Nov;6(11):1033-8.

A computational approach to simplifying the protein folding alphabet.

Author information

  • 1National Laboratory of Solid-State Microstructure and Department of Physics, Nanjing University, Nanjing 210093, People's Republic of China.

Abstract

What is the minimal number of residue types required to form a structured protein? This question is important for understanding protein modeling and design. Recently, an experimental finding by Baker and coworkers suggested a five-residue solution to this problem. We were motivated by their results and by the arguments of Wolynes to study reductions of protein representation based on the concept of mismatch between a reduced interaction matrix and the Miyazawa and Jernigan (MJ) matrix. We find several possible simplified schemes from the relationship of minimized mismatch versus the number of residue types (N = approximately 2-20). As a specific case, an optimal reduction with five types of residues has the same form as the simplified palette of Baker and coworkers. Statistical and kinetic features of a number of sequences are tested. Comparison of results from sequences with 20 residue types and their reduced representations indicates that the reduction by mismatch minimization is successful. For example, sequences with five types of residues have good folding ability and kinetic accessibility in model studies.

Comment in

PMID:
10542095
DOI:
10.1038/14918
[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Nature Publishing Group
    Loading ...
    Support Center