Controlling for P-value inflation in allele frequency change in experimental evolution and artificial selection...

Area of Science:

• Evolutionary biology
• Genetics
• Population genetics

Background:

• Experimental evolution studies identify genomic regions under selection by detecting allele frequency changes.
• Current methods may overestimate loci under selection due to P-value inflation, potentially caused by population stratification like relatedness.
• Genomewide association (GWA) studies offer tools to address such inflation.

Purpose of the Study:

• To adapt GWA methods for detecting allele frequency changes in experimental evolution.
• To account for population relatedness and reduce false positives in selection studies.
• To investigate the impact of selection on eco-evolutionary dynamics.

Main Methods:

• Utilized GWA methods by treating individual presence post-selection as a binary response variable.
• Applied methods to simulated data with varying population structures.
• Validated findings using empirical data from an artificial selection experiment on house sparrows (Passer domesticus) for tarsus length.
• Employed genomic control to correct for P-value inflation in empirical data.

Main Results:

• The proposed GWA approach effectively reduces false positives in allele frequency change tests when accounting for relatedness.
• Detecting causal loci under selection showed reduced power after accounting for relatedness.
• Empirical data confirmed P-value inflation in allele frequency changes.
• Genomic control successfully corrected for inflation in empirical data.

Conclusions:

• Accounting for relatedness is crucial for accurate estimation of allele frequency change in experimental evolution.
• Selection on heritable traits can lead to widespread and immediate genomic allele frequency shifts.
• These shifts have significant implications for population eco-evolutionary dynamics.