Exact limits of inference in coalescent models

Area of Science:

• Population Genetics
• Computational Biology
• Evolutionary Biology

Background:

• Inferring population size history from molecular data is crucial in population genetics.
• Coalescent models link population size to genealogies but are computationally intensive.
• Current methods often use data subsets or summary statistics, limiting accuracy.

Purpose of the Study:

• To develop methods for accurately recovering population size history from molecular sequence data.
• To provide exact probabilities for distinguishing between alternative population size histories.
• To extend analyses to models with recombination.

Main Methods:

• Utilized coalescent time data to analyze population size history.
• Derived exact expressions for hypothesis distinction probabilities.
• Employed Monte Carlo simulations for complex scenarios and recombination models.

Main Results:

• Developed exact formulas for distinguishing population size histories based on data parameters.
• Demonstrated more conservative inferential limits compared to prior research.
• Extended the analysis to incorporate pairwise SMC and SMC' recombination models.

Conclusions:

• The derived methods offer precise probabilities for inferring population size history.
• Results suggest stricter limitations on accurate inference than previously understood.
• This work aids in designing studies for testing hypotheses on population size dynamics.