Performance Improvement of Genetic Algorithm for Multiple Sequence Alignment

Abstract

The multiple sequence alignment (MSA) is considered one of the most important tasks in Bioinformatics. Nevertheless, with the growth in the amount of genomic data available, it is essential the results with biological significance and an acceptable execution time. Thus, many tools have been proposed with the focus in these two last requirements. Considering the tools, the MSA-GA is of them, which is based on Genetic Algorithms approach, and it is widely used to perform MSA, because its simpler approach and good results. However, the biological significance and execution time are two elements that work in opposite directions, because when more biological significance is desired, more execution time will be wasted, mainly considering the amount of genomic data produced by next generation sequencing recently. Therefore, the implementation of parallel programming can help to smooth this disadvantage. Thus, in the present work we developed a parallel version of the MSA-GA tool using multithread programming, in order to keep the good results produced by the tool and improving its execution time.