Using a Bio-Inspired Algorithm to Resolve the Multiple Sequence Alignment Problem

One of the most challenging tasks in bioinformatics is the resolution of Multiple Sequence Alignment (MSA) problem. It consists in comparing a set of protein or DNA sequences, in aim of predicting their structure and function. This paper introduces a new bio-inspired approach to solve such problem....

Full description

Saved in:
Bibliographic Details
Published inInternational journal of applied metaheuristic computing Vol. 7; no. 3; pp. 36 - 55
Main Authors Zemali, El-amine, Boukra, Abdelmadjid
Format Journal Article
LanguageEnglish
Published Hershey IGI Global 01.07.2016
Subjects
Algorithms
Alignment
Analysis
Bioinformatics
Convergence
DNA sequencing
Evolutionary algorithms
Gene sequencing
Nucleotide sequencing
Online AccessGet full text

Cover

More Information
Summary:One of the most challenging tasks in bioinformatics is the resolution of Multiple Sequence Alignment (MSA) problem. It consists in comparing a set of protein or DNA sequences, in aim of predicting their structure and function. This paper introduces a new bio-inspired approach to solve such problem. This approach named BA-MSA is based on Bat Algorithm. Bat Algorithm (BA) is a recent evolutionary algorithm inspired from Bats behavior seeking their prey. The proposed approach includes new mechanism to generate initial population. It consists in generating a guide tree for each solution with progressive approach by varying some parameters. The generated guide tree will be enhanced by Hill-Climbing algorithm. In addition, to deal with the premature convergence of BA, a new restart technique is proposed to introduce more diversification when detecting premature convergence. Balibase 2.0 datasets are used for experiments. The comparison with well-known methods as MSA-GA MSA-GA (w\prealign), ClustalW, and SAGA and recent method (BBOMP) shows the effectiveness of the proposed approach.
ISSN:1947-8283
1947-8291
DOI:10.4018/IJAMC.2016070103