Conditional entropy in variation-adjusted windows detects selection signatures associated with expression...

Area of Science:

• Population Genetics
• Human Evolution
• Genomics

Background:

• Human genetic variation has been shaped by evolutionary pressures over 50,000 years.
• Environmental and social interactions drive genetic differences, influencing modern health and disease.
• Regulatory variants, key to gene expression, are investigated for signs of positive selection.

Purpose of the Study:

• To develop and validate a novel method for detecting positive selection in genetic variants.
• To specifically test if regulatory variants (cis-eQTLs) are under positive selection.
• To compare the efficacy of the new method against existing positive selection detection approaches.

Main Methods:

• Introduced Adjusted Haplotype Conditional Entropy (H|H) to detect recent local selection signals.
• Applied H|H and existing methods to cis-eQTLs from six genotype-expression studies.
• Utilized conditional logistic regression, adjusted for Minor Allele Frequency (MAF), to compare selection signals.

Main Results:

• H|H demonstrated significantly higher power in identifying cis-eQTLs under positive selection compared to existing methods.
• Strong positive selection signals were detected in specific Gene Ontology (GO) categories, particularly for regulatory variants.
• Some cis-eQTLs, notably in innate immunity, showed weak or no detectable positive selection signatures, suggesting alternative evolutionary pressures.

Conclusions:

• The H|H method provides a robust signal for identifying cis-eQTLs under positive selection.
• Positive selection on cis-eQTLs is evident in certain biological processes relevant to human health.
• The findings suggest that different modes of selection (e.g., balancing selection) may act on regulatory variants in other biological contexts.