Survey of the use of genetic algorithm for multiple sequence alignment

Abstract

Multiple Sequence Alignment (MSA) is used in genomic analysis, such as the identification of conserved sequence motifs, the estimation of evolutionary divergence between sequences, and the genes’ historical relationships inference. Several researches were conducted to determine the level of similarity of a set of sequences. Due to the problem of the NP-complete class property, a number of researches use genetic algorithms (GA) to find a solution to the multiple sequence alignment. However, the nature of genetic algorithms makes the complexity extremely high due to the redundancy provided by the different operators. The aim of this paper is to study some proposed GA solutions provided for MSA and to compare them using some criteria which we believe any solution should comply with in matters of representativeness, closeness and original sequence invariance.