Renewing Felsenstein’s Phylogenetic Bootstrap in the Era of Big Data

Abstract

Felsenstein’s article describing the application of the bootstrap to evolutionary trees is one of the most cited papers of all time. The bootstrap method, based on resampling and replications, is used extensively to assess the robustness of phylogenetic inferences. However, increasing numbers of sequences are now available for a wide variety of species, and phylogenies with hundreds or thousands of taxa are becoming routine. In that framework, Felsenstein’s bootstrap tends to yield very low supports, especially on deep branches. We propose a new version of phylogenetic bootstrap, in which the presence of inferred branches in replications is measured using a gradual “transfer” distance, as opposed to the original version using a binary presence/absence index. The resulting supports are higher, while not inducing falsely supported branches. Our method is applied to large mammal, HIV, and simulated datasets, for which it reveals the phylogenetic signal, while Felsenstein’s bootstrap fails to do so.