The Brain Storm Dynamic Technique DBSO-MSA for Efficiently Resolving Multiple Sequence Alignment

Abstract

Multiple Sequence Alignment (MSA) is a critical step in molecular biology. Different techniques are having been proposed for obtaining optimal alignments, still, there is a need of developing accurate and efficient techniques for optimal sequence alignment. One of the efficient techniques among the swarm optimization families is Brain Storm Optimization Technique based on human social behavior has achieved success in numerous applications. However, population divergence plays a major role in obtaining better solutions for optimization problems. Therefore, high diverged populations obtain optimal results. The multiple sequences alignment is an efficient optimization for dataset analysis but hidden samples do not get tracked by MSA. Therefore DBSO_MSA model requirement is there to crossover limitations of the above model. This paper proposed a dynamic clustered and populated Brain Storm Optimization Algorithm for obtaining more optimal alignment solutions for the Multiple Sequence Alignment problem (DBSO-MSA). The dynamic nature with respect to the number of clusters and population generation at every iteration is incorporated into BSO. The number of solutions and cluster size at each iteration is controlled by the probability variable either it increases or decreases the solution space to explore more diversification in obtaining alignments for the MSA problem. The experiments show DBSO-MSA effectively improves the alignment score on the benchmark sequence datasets compared to the Classical BSO and other evolutionary algorithms.