Detecting loci under selection in a hierarchically structured population.

Area of Science:

• Population Genetics
• Evolutionary Biology
• Genomics

Background:

• Genetic diversity patterns reveal loci under selection, with high F(ST) for local adaptation and low F(ST) for balancing selection.
• Standard F(ST) tests often use a simple island model, which may not accurately reflect complex population structures like recent shared ancestry or gene flow barriers.

Purpose of the Study:

• To propose and validate a hierarchical island model for generating genetic diversity distributions in complex population structures.
• To assess the impact of hierarchical structure on tests for selection and reduce false positive findings in genome scans.

Main Methods:

• Developed a hierarchical island model accounting for differential migration within and between population groups.
• Generated joint distributions of genetic diversity within and between populations.
• Incorporated the mutational process, crucial for short tandem repeat (STR) data.

Main Results:

• Tests ignoring hierarchical structure produce a significant excess of false positive loci when such structure exists.
• The proposed hierarchical island model is robust to uncertainties in group and deme numbers.
• Application to human and stickleback STR data yielded fewer significant loci compared to non-hierarchical models, indicating reduced false positives.

Conclusions:

• Accounting for hierarchical population structure is crucial for accurate detection of selection.
• The hierarchical island model provides a more robust framework for genome scans, especially with STR data.
• This approach enhances the precision of identifying gene classes under selection.