Variational phylogenetic inference with products over bipartitions

TL;DR

This paper introduces a novel variational Bayesian method for ultrametric phylogenetic trees that performs inference over all tree space without MCMC, offering efficiency and competitive accuracy.

Contribution

It develops a new variational family based on coalescent times and derives a closed-form density, enabling efficient inference over all tree space for ultrametric trees.

Findings

Achieves competitive accuracy on genomic datasets

Requires fewer gradient evaluations than existing methods

Successfully applied to SARS-CoV-2 viral RNA data

Abstract

Bayesian phylogenetics is vital for understanding evolutionary dynamics, and requires accurate and efficient approximation of posterior distributions over trees. In this work, we develop a variational Bayesian approach for ultrametric phylogenetic trees. We present a novel variational family based on coalescent times of a single-linkage clustering and derive a closed-form density for the resulting distribution over trees. Unlike existing methods for ultrametric trees, our method performs inference over all of tree space, it does not require any Markov chain Monte Carlo subroutines, and our variational family is differentiable. Through experiments on benchmark genomic datasets and an application to the viral RNA of SARS-CoV-2, we demonstrate that our method achieves competitive accuracy while requiring significantly fewer gradient evaluations than existing state-of-the-art techniques.