A PSO with Improved Initialization Operator for Solving Multiple Sequence Alignment Problems

Abstract

In this paper, an improved particle swarm optimization (PSO) technique is explored for aligning multiple sequences. PSO has recently emerged as a new randomized heuristic method for both real-valued and discrete optimization problems. This is a nature-inspired algorithm based on the movement and intelligence of swarms. Here each solution is represented in encoded form as ‘position’ like ‘chromosome’ in genetic algorithm (GA). The fitness function is designed accordingly to optimize the objective functions, i.e., maximizing the matching components of the sequences and reducing the number of mismatched components in the sequences. The performance of the proposed method has been tested on publicly available benchmark datasets (i.e., Bali base) to establish the potential of PSO to solve alignment problem with better and/or competitive performance. The results are compared with some of the well known existing methods available in literature such as DIALIGN, HMMT, ML-PIMA PILEUP8, and RBT-GA. The experimental results showed that proposed method attained better solutions than the others for most of the cases.