Resolving competing evolutionary histories in joint ancestral state reconstruction.

Area of Science:

• Comparative biology
• Evolutionary biology
• Phylogenetics

Background:

• Ancestral state reconstruction (ASR) is crucial for understanding evolutionary history.
• Current ASR methods predominantly use marginal reconstructions, focusing on individual nodes.
• Marginal reconstructions are suitable for node-specific hypotheses but not for identifying evolutionary trajectories.

Purpose of the Study:

• To introduce and validate joint reconstruction methods for characterizing evolutionary trajectories.
• To provide tools for quantifying and summarizing uncertainty in joint ancestral histories.
• To demonstrate the utility of joint reconstructions for complex evolutionary scenarios.

Main Methods:

• Utilized conditional probabilities derived from stochastic mapping to sample plausible ancestral histories.
• Developed and applied tools for quantifying and summarizing joint uncertainty.
• Employed simulations and an empirical case study for validation.

Main Results:

• Joint reconstructions more effectively recover simulated trait histories compared to marginal estimates.
• Uncertainty surrounding ancestral histories derived from joint reconstructions is biologically meaningful.
• Analysis of multidrug-resistant *Klebsiella pneumoniae* revealed multiple competing histories of antibiotic resistance evolution.

Conclusions:

• Joint reconstruction provides a more comprehensive approach to understanding evolutionary trajectories than marginal methods.
• The study highlights the complexity of antibiotic resistance evolution, characterized by multiple distinct phenotype-genotype transition histories.
• The developed methods offer critical implications for predicting and understanding pathogen evolution.