Maximum likelihood and Bayesian methods for estimating the distribution of selective effects among classes of...

Area of Science:

• Population Genetics
• Statistical Genomics
• Evolutionary Biology

Background:

• Understanding natural selection on DNA polymorphism is crucial for evolutionary studies.
• Hierarchical statistical models offer a powerful framework for analyzing complex genetic data.
• Previous methods may not adequately capture selection across diverse mutation classes.

Purpose of the Study:

• To present maximum likelihood and Bayesian approaches for analyzing natural selection on DNA polymorphism.
• To develop computational methods for estimating selection intensity across the genome.
• To provide a framework for analyzing natural selection on different classes of single nucleotide polymorphisms (SNPs).

Main Methods:

• Bayesian analysis using Markov chain Monte-Carlo (MCMC) for posterior distribution sampling.
• Frequentist analysis employing an Expectation-Maximization (EM) algorithm for maximum likelihood estimation.
• Development of a statistical framework for hierarchical models of natural selection.

Main Results:

• Efficient algorithms for both Bayesian and frequentist analyses of natural selection models.
• Methods to estimate genome-wide mean and variance in selection intensity among mutation classes.
• A flexible framework applicable to dispersed mutations and SNP analysis.

Conclusions:

• The presented statistical framework effectively models natural selection on DNA polymorphism.
• The developed MCMC and EM algorithms provide robust tools for genetic data analysis.
• This approach enhances the understanding of how natural selection shapes genetic variation across different mutation types.