A genetic algorithm for alignment of multiple DNA sequences

Abstract

This paper presents a new genetic algorithm based solution to obtain alignment of multiple DNA molecular sequences. Multiple Sequence alignment is one of the most active ongoing research problems in the field of computational molecular biology. Sequence alignment is important because it allows scientists to analyze protein strands (such as DNA and RNA) and determine where there are overlaps. These overlaps can show commonalities in evolution and they also allow scientists to better prepare vaccines against viruses, which are made of protein strands. We have proposed new genetic operations for crossover, mutation, fitness calculation, population initialization. Proposed scheme generates new populations with better fitness value. We have also reviewed the some of the popular works by different researchers towards solving the MSA problem w.r.t various phases involved in general GA procedure. A working example is presented to validate the proposed scheme. Improvement in the overall population fitness is also calculated. © 2012 ICST Institute for Computer Science, Social Informatics and Telecommunications Engineering.