Phylogeny Estimation by Integration over Isolation with Migration Models.

Area of Science:

• Evolutionary biology
• Population genetics
• Bioinformatics

Background:

• Phylogenetic inference is complex for closely related groups, particularly with historical gene flow.
• Existing methods often struggle to fully account for migration and genetic drift simultaneously.

Purpose of the Study:

• To develop a novel statistical framework for robust phylogeny estimation in the presence of gene flow.
• To apply this method to understand the evolutionary history of human and chimpanzee populations.

Main Methods:

• A hierarchical Bayesian, Markov-chain Monte Carlo (MCMC) approach was implemented.
• The method utilizes a 'hidden genealogy' augmentation for efficient phylogenetic updating.
• It operates within a full Isolation-with-Migration model, incorporating directional gene flow and genetic drift.

Main Results:

• Analysis of African hunter-gatherer data supported a clear phylogenetic history with significant gene exchange with an unsampled 'ghost' population.
• Ancient divergence between this ghost population and modern humans was inferred, aligning with human/archaic divergence hypotheses.
• Chimpanzee population data yielded a clear phylogeny, indicating DNA exchange between several pairs but no evidence for an unsampled ghost population.

Conclusions:

• The developed method provides a powerful, likelihood-based tool for inferring species or population phylogenies under complex demographic scenarios.
• It offers new insights into human evolutionary history, suggesting ancient admixture events.
• The findings highlight differences in evolutionary histories between human and chimpanzee populations.