Evolutionary algorithms for supertree search

Phylogenetic inference of the history of life on Earth continues to be a major effort of evolutionary biology. Such inference can be accomplished through the use of individual genes, sets of genes, or complete genomes. While the latter may provide the most robust description of the true phylogenetic history, the computational demands of complete genome comparison and phylogenetic construction is daunting. Thus most researchers are left using sets of conserved genes for the resolution of a common phylogeny (what is termed a “supertree” search). However as the number of taxa increases or as the number of source trees used in construction of a supertree increases, the number of possible supertree solutions increases tremendously. This requires consideration of alternate methods to search this space efficiently such as those that use stochastic methods. Here for the first time we present a method for supertree search using evolutionary algorithms and evaluate its utility on a set of derived supertree problems with 50 taxa. The results indicate the utility of this approach and offer opportunities for future refinement.