BAYESFST

BAYESFST estimates the coancestry coefficient F_ST using a hierarchical Bayesian framework with Markov chain Monte Carlo (MCMC) inference to detect loci under natural selection.

Key Features:

• Hierarchical Bayesian Framework: Employs a hierarchical Bayesian model to estimate F_ST while incorporating prior information and parameter uncertainty.
• Markov Chain Monte Carlo (MCMC): Implements MCMC methods to explore complex posterior distributions of genetic parameters.
• Simulation-Based Validation: Validated using simulations of structured populations with numerous unlinked, diallelic loci including both neutral and selected loci.
• Comparison with Frequentist Methods: Compares Bayesian estimates to moment-based frequentist estimates of F_ST for identifying loci under selection.
• Sensitivity to Selection vs. Migration: Distinguishes loci effectively when the selection coefficient is substantially higher than the migration rate and shows comparable detection to frequentist methods when the selection coefficient is at least five times the migration rate.
• Limitation for Balancing Selection: Shows limited ability to reliably distinguish loci under balancing selection even with high selection coefficients.

Scientific Applications:

• Detecting Signatures of Natural Selection: Applied in genomic surveys to detect loci subject to natural selection.
• Identifying Functionally Important Loci: Pinpoints candidate loci that may be involved in disease causation.
• Studying Adaptive Differentiation: Investigates adaptive differentiation among populations under selective pressures.
• Investigating Speciation Hypotheses: Analyzes allele frequency differences and loci under selection to inform speciation studies.

Methodology:

Simulations of structured populations with a mix of neutral and selected unlinked, diallelic loci are analyzed using a hierarchical Bayesian model with MCMC inference; results are compared to moment-based frequentist F_ST estimates, and detection performance is reported relative to the ratio of selection coefficient to migration rate, including noted limitations for balancing selection.