A Minimal yet Flexible Likelihood Framework to Assess Correlated Evolution

Area of Science:

• Evolutionary biology
• Phylogenetics
• Computational biology

Background:

• Understanding evolutionary processes requires analyzing trait changes over time.
• Characterizing correlated evolution, where multiple evolutionary processes interact, is crucial but computationally challenging with existing methods.

Purpose of the Study:

• To develop a minimal likelihood framework for modeling the joint evolution of two discrete traits on a phylogenetic tree.
• To provide a computationally efficient method for studying correlated evolution.

Main Methods:

• A minimal likelihood framework was developed to model the joint evolution of two traits.
• The framework uses a few parameters to tune mutation rates and their interdependencies.
• The method was validated using simulations and 16S rRNA sequence data from enterobacteria.

Main Results:

• The proposed framework efficiently characterizes the type and strength of correlated evolution.
• It allows for testing independence, identifying interaction types, and estimating model parameters.
• The method demonstrated good performance in parameter estimation and model selection, even with small datasets (<100 species).

Conclusions:

• The new framework offers an efficient and powerful approach to study correlated evolution.
• It advances the understanding of how multiple evolutionary processes interact.
• The method is applicable to various discrete traits in evolutionary biology.