An Optimization Approach for Multiple Sequence Alignment using Divide-Conquer and Genetic Algorithm

Abstract

Multiple Sequence Alignment (MSA) is a very effective tool in bioinformatics. It is used for the prediction of the structure and function of the protein, locating DNA regulatory elements like binding sites, and evolutionary analysis. This research paper proposed an optimization method for the improvement of multiple sequence alignment generated through progressive alignment. This optimization method consists of a fusion of two problem-solving techniques, divide-conquer and genetic algorithms in which the initial population of MSAs was generated through progressive alignment. Each multiple alignment was then divided vertically into four parts, three genetic operators were applied on each part of the MSA, recombination was done to reconstruct the full MSA. To estimate the performance of the method the results generated through the method are compared with well-known existing MSA methods named Clustal Ω, MUSCLE, PRANK, and Clustal W. Experimental results showed an 11-26% increase in sum_of_pair score (SP score) in the proposed method in comparison to the above-mentioned methods. SP score is the cumulative score of all possible pairs of alignment within the MSA.