A phylogenomics approach for selecting robust sets of phylogenetic markers

Reconstructing the evolutionary relationships of species is a major goal in biology. Despite the increasing number of completely sequenced genomes, a large number of phylogenetic projects rely on targeted sequencing and analysis of a relatively small sample of marker genes. The selection of these phylogenetic markers should ideally be based on accurate predictions of their combined, rather than individual, potential to accurately resolve the phylogeny of interest. Here we present and validate a new phylogenomics strategy to efficiently select a minimal set of stable markers able to reconstruct the underlying species phylogeny. In contrast to previous approaches, our methodology does not only rely on the ability of individual genes to reconstruct a known phylogeny, but it also explores the combined power of sets of concatenated genes to accurately infer phylogenetic relationships of species not previously analyzed. We applied our approach to two broad sets of cyanobacterial and ascomycetous fungal species, and provide two minimal sets of six and four genes, respectively, necessary to fully resolve the target phylogenies. This approach paves the way for the informed selection of phylogenetic markers in the effort of reconstructing the tree of life.