Home // BIOTECHNO 2011, The Third International Conference on Bioinformatics, Biocomputational Systems and Biotechnologies // View article
Authors:
Hazem R. Ahmed
Janice I. Glasgow
Keywords: protein structure prediction; protien contact maps; structural pattern matching; evolutionary information; harrell-davis quantiles.
Abstract:
Protein structure prediction from the primary sequence remains a major challenging problem in bioinformatics. The main issue here is that it is computationally complex to reliably predict the full three-dimensional structure of a protein from its one-dimensional sequence. A two-dimensional contact map has, therefore, been used as an intermediate step in this problem. A contact map is a simpler, yet representative, alternative for the three-dimensional protein structure. In this paper, we propose a pattern matching approach to locate similar substructural patterns between protein contact map pairs using protein sequence information. These substructural patterns are of particular interest to our research, because they could ultimately be used as building blocks for a bottom-up approach to protein structure prediction from contact maps. We further demonstrate how to improve the performance of identifying these patterns by incorporating both protein sequence and evolutionary information. The results are benchmarked using a large standard protein dataset. We performed statistical analyses (e.g., Harrell-Davis Quantiles and Bagplots) that show sequence information is more helpful in locating short-range contacts than long-range contacts. Moreover, incorporating evolutionary information has remarkably improved the performance of locating similar short-range contacts between contact map pairs.
Pages: 49 to 55
Copyright: Copyright (c) IARIA, 2011
Publication date: May 22, 2011
Published in: conference
ISSN: 2308-4383
ISBN: 978-1-61208-137-3
Location: Venice/Mestre, Italy
Dates: from May 22, 2011 to May 27, 2011