An Efficient Bi-Level Discrete PSO Variant for Multiple Sequence Alignment

Abstract

This paper implements bi-level discrete particle swarm optimization (BL-DPSO), an efficient discrete PSO variant for multiple sequence alignment (MSA) problem of nucleotides. Level one works on optimizing dimension for entire swarm, i.e., obtaining the optimal sequence length, whereas level two works for optimizing each and every particles position, i.e., to attain optimum gap positions for maximum alignment score. Set-theory based position and velocity update strategies are implemented in the proposed BL-DPSO. The capability of the proposed approach is evaluated with three standard scoring schemes at specific parameters with two types of benchmark datasets of DNA and RNA. BL-DPSO alignments are compared with four PSO variants, i.e., S-PSO, M-PSO, CPSO-S$_k$, and TL-PSO, and two leading alignment software, i.e., ClustalW and T-Coffee, at different alignment scores. Obtained results prove the competency of BL-DPSO at accuracy aspects.