Hybrid genetic algorithms for the determination of DNA motifs to satisfy postulate 2-Optimality

Abstract

Currently, determining DNA motifs or consensus plays an indispensable role in bioinformatics. Many postulates have been proposed for finding a consensus. Postulate 2-Optimality is essential for this task. A consensus satisfying postulate 2-Optimality is the best representative of a profile, and its distances to the profile members are uniform. However, this postulate has not been widely investigated in identifying a DNA motif or consensus for a DNA motif profile. The HDC algorithm is the best at this task in the literature. This study focuses on determining DNA motifs that satisfy postulate 2-Optimality. We propose a new hybrid genetic (HG1) algorithm based on the elitism strategy and local search. Subsequently, a novel elitism strategy and longest distance strategy are introduced to maintain the balance of exploration and exploitation. A new hybrid genetic (HG2) algorithm is developed based on the proposed exploration and exploitation balance approach. The simulation results show that these algorithms provide a high-quality DNA motif. The HG2 algorithm provides a DNA motif with the best quality.