A genome-scan method to identify selected loci appropriate for both dominant and codominant markers: a Bayesian...

Area of Science:

• Population genetics
• Evolutionary biology
• Genomics

Background:

• Identifying loci under natural selection is crucial in evolutionary and genomic research.
• Existing methods often rely on population differentiation coefficients to detect outliers, separating neutral from adaptive effects.

Purpose of the Study:

• To extend existing outlier detection methods by developing a Bayesian approach to directly estimate the probability of selection at each locus.
• To adapt the method for dominant markers, such as Amplified Fragment Length Polymorphisms (AFLPs).

Main Methods:

• A Bayesian statistical framework was employed to estimate locus-specific selection probabilities.
• The model was extended to accommodate dominant markers (e.g., AFLPs) alongside codominant markers (SNPs, microsatellites).
• Sensitivity analyses were performed using simulated and real data, including human and Littorina saxatilis datasets.

Main Results:

• The Bayesian method provides a direct estimate of selection probability per locus.
• The model demonstrates robustness to complex demographic histories for neutral genetic differentiation.
• Inclusion of isolated, bottlenecked populations can inflate false positive rates, necessitating careful population selection for analysis.

Conclusions:

• The developed Bayesian method offers a powerful tool for identifying loci under natural selection.
• Careful consideration of population structure, particularly bottlenecked populations, is essential to ensure accurate results.
• The method is versatile, applicable to various marker types and demographic scenarios, and implemented in accessible software.