A Parallel Multiobjective Algorithm Inspired by Fireflies for Inferring Evolutionary Trees on Multicore Machines

摘要

Recent researches have pointed out the need to combine parallelism and bioinspired computing to address computationally intensive problems in bioinformatics. The inference of evolutionary histories represents one of the most complex problems in this field. Phylogenetic inference can be tackled by using multiobjective metaheuristics designed to resolve the problems that arise when different optimality criteria support conflicting evolutionary relationships. As the inference process becomes harder when we have to consider multiple criteria simultaneously, these new approaches must be defined on the basis of parallel computing. In this paper, we propose a parallel multiobjective approach inspired by fireflies to address the phylogenetic inference problem by using OpenMP to exploit the characteristics of multicore machines. Experimental results on four real biological data sets show significant parallel and biological performances with regard to other proposals from the literature.