By Topic

Enhancement of Initial Equivalency for Protein Structure Alignment Based on Encoded Local Structures

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

The purchase and pricing options are temporarily unavailable. Please try again later.
6 Author(s)
Hung, K. ; Inst. of Biomed. Eng., Nat. Taiwan Univ., Taipei, Taiwan ; Jui-Chih Wang ; Cheng-Wei Chen ; Cheng-Long Chuang
more authors

Most alignment algorithms find an initial equivalent residue pair followed by an iterative optimization process to explore better near-optimal alignments in the surrounding solution space of the initial alignment. It plays a decisive role in determining the alignment quality since a poor initial alignment may make the final alignment trapped in an undesirable local optimum even with an iterative optimization. We proposed a vector-based alignment algorithm with a new initial alignment approach accounting for local structure features called MIRAGE-align. The new idea is to enhance the quality of the initial alignment based on encoded local structural alphabets to identify the protein structure pair whose sequence identity falls in or below twilight zone. The statistical analysis of alignment quality based on match index and computation time demonstrated that MIRAGE-align algorithm outperformed four previously published algorithms, i.e., the residue-based algorithm, the vector-based algorithm, TM-align, and Fr-TM-align. MIRAGE-align yields a better estimate of initial solution to enhance the quality of initial alignment and enable the employment of a noniterative optimization process to achieve a better alignment.

Published in:

Information Technology in Biomedicine, IEEE Transactions on  (Volume:16 ,  Issue: 6 )