Improvement of protein structure comparison using a structural alphabet

Joseph, Agnel Praveen and Srinivasan, N and de Brevern, Alexandre G (2011) Improvement of protein structure comparison using a structural alphabet. In: Biochimie, 93 (9). pp. 1434-1445.

PDF Improvement_of.pdf - Published Version Restricted to Registered users only Download (1MB) | Request a copy

Abstract

The three dimensional structure of a protein provides major insights into its function. Protein structure comparison has implications in functional and evolutionary studies. A structural alphabet (SA) is a library of local protein structure prototypes that can abstract every part of protein main chain conformation. Protein Blocks (PBS) is a widely used SA, composed of 16 prototypes, each representing a pentapeptide backbone conformation defined in terms of dihedral angles. Through this description, the 3D structural information can be translated into a 1D sequence of PBs. In a previous study, we have used this approach to compare protein structures encoded in terms of PBs. A classical sequence alignment procedure based on dynamic programming was used, with a dedicated PB Substitution Matrix (SM). PB-based pairwise structural alignment method gave an excellent performance, when compared to other established methods for mining. In this study, we have (i) refined the SMs and (ii) improved the Protein Block Alignment methodology (named as iPBA). The SM was normalized in regards to sequence and structural similarity. Alignment of protein structures often involves similar structural regions separated by dissimilar stretches. A dynamic programming algorithm that weighs these local similar stretches has been designed. Amino acid substitutions scores were also coupled linearly with the PB substitutions. iPBA improves (i) the mining efficiency rate by 6.8% and (ii) more than 82% of the alignments have a better quality. A higher efficiency in aligning multi-domain proteins could be also demonstrated. The quality of alignment is better than DALI and MUSTANG in 81.3% of the cases. Thus our study has resulted in an impressive improvement in the quality of protein structural alignment. (C) 2011 Elsevier Masson SAS. All rights reserved.

Item Type:	Journal Article
Publication:	Biochimie
Publisher:	Elsevier Science
Additional Information:	Copyright of this article belongs to Elsevier Science.
Keywords:	Amino acid;Structural alphabet;Protein blocks;Semi-global alignment;Anchor-based alignment;Protein folds;Structural comparison;Protein structure mining;Protein data bank
Department/Centre:	Division of Biological Sciences > Molecular Biophysics Unit
Date Deposited:	16 Sep 2011 05:17
Last Modified:	16 Sep 2011 05:17
URI:	http://eprints.iisc.ac.in/id/eprint/40435

Actions (login required)

View Item