BayeScan

BayeScan detects candidate loci under natural selection by comparing allele frequencies between populations using a multinomial-Dirichlet Bayesian model.

Key Features:

• Bayesian F(ST) outlier approach: Implements a Bayesian F(ST) outlier framework for identifying loci with allele frequency differentiation consistent with selection.
• Multinomial-Dirichlet model: Employs a multinomial-Dirichlet statistical model to model allele frequency distributions across populations.
• AFLP-specific analysis: Tailored for amplified fragment length polymorphism (AFLP) data and accounts for AFLP band intensity rather than only presence/absence.
• Allele frequency and F(IS) estimation: Produces population-based estimates of allele frequencies and fixation index (F(IS)) values from AFLP signal intensities.
• Enhanced statistical power and specificity: Demonstrates increased power to detect AFLP markers under selection approaching the efficacy of SNP data while maintaining high specificity.
• Genome-scan applicability: Enables genome scans for candidate loci without prior knowledge of specific variants or traits under selection.

Scientific Applications:

• Local adaptation scans: Used to uncover genetic bases of local adaptation, including detecting loci associated with environmental contrasts such as high- versus low-altitude populations.
• Empirical study in Microtus arvalis: Applied to 3,027 AFLP markers across four populations of the common vole (Microtus arvalis), identifying 20 candidate markers showing significant allele frequency differences between ~2000 m and <600 m altitude populations.

Methodology:

Performs a Bayesian F(ST) outlier analysis using a multinomial-Dirichlet model and leverages AFLP band intensity to estimate allele frequencies and F(IS).