Applying a multiobjective metaheuristic inspired by honey bees to phylogenetic inference

On the design of shared memory approaches to parallelize a multiobjective bee-inspired proposal for phylogenetic reconstruction

“…In a previous research, this algorithm was applied to tackle the problem by performing direct searches over the tree space, using an individual representation based on tree-shaped data structures [45]. In this work, we propose a new implementation based on a distance-based methodology which has shown significant results when compared to other classical approaches [41].…”

“…This reasoning was justified in [45], where multiobjective techniques were successfully applied to solve a well-known dataset of Plethodontid salamander DNA [35]. The analysed multiobjective tree was able to provide a satisfying phylogenetic explanation to the evolution of these organisms, solving those clades where parsimony and likelihood showed real conflict [12,53].…”

“…The analysed multiobjective tree was able to provide a satisfying phylogenetic explanation to the evolution of these organisms, solving those clades where parsimony and likelihood showed real conflict [12,53]. More details about the underlying trade-off between these criteria, along with the comparison of single and multiobjective topologies can be found in [45].…”

“…This issue motivated Cancino et al to develop MPI and MPI/OpenMP schemes for PhyloMOEA, a tool for inferring phylogenies under parsimony and likelihood [9]. In 2013, Santander-Jiménez and Vega-Rodríguez [45] studied the biological implications of multiobjective optimization by applying a novel metaheuristic inspired by honey bees with tree-shaped individual representation. The results obtained by this first version of MOABC showed a significant improvement over the standard NSGA-II in terms of multiobjective and biological performance.…”

“…For this purpose, we study shared memory schemes based on OpenMP [10] to parallelize a Multiobjective Artificial Bee Colony (MOABC) algorithm for phylogenetics [45]. In order to exploit pure shared memory environments, two parallel approaches are proposed.…”