Improved Variational Bayesian Phylogenetic Inference using Mixtures

TL;DR

VBPI-Mixtures significantly improves phylogenetic posterior inference by effectively modeling complex tree-topology distributions using mixture learning within a variational Bayesian framework, outperforming previous methods.

Contribution

Introduces VBPI-Mixtures, a novel algorithm that enhances tree-topology posterior approximation in phylogenetics through mixture learning, addressing limitations of existing variational methods.

Findings

Achieves state-of-the-art density estimation on real datasets.

Successfully models multimodal tree-topology distributions.

Outperforms existing methods in accuracy and robustness.

Abstract

We present VBPI-Mixtures, an algorithm designed to enhance the accuracy of phylogenetic posterior distributions, particularly for tree-topology and branch-length approximations. Despite the Variational Bayesian Phylogenetic Inference (VBPI), a leading-edge black-box variational inference (BBVI) framework, achieving remarkable approximations of these distributions, the multimodality of the tree-topology posterior presents a formidable challenge to sampling-based learning techniques such as BBVI. Advanced deep learning methodologies such as normalizing flows and graph neural networks have been explored to refine the branch-length posterior approximation, yet efforts to ameliorate the posterior approximation over tree topologies have been lacking. Our novel VBPI-Mixtures algorithm bridges this gap by harnessing the latest breakthroughs in mixture learning within the BBVI domain. As a result, VBPI-Mixtures is capable of capturing distributions over tree-topologies that VBPI fails to model. We deliver state-of-the-art performance on difficult density estimation tasks across numerous real phylogenetic datasets.