Detecting recent positive selection with high accuracy and reliability by conditional coalescent tree.

Area of Science:

• Human evolutionary genetics
• Population genetics
• Genomics

Background:

• Understanding human evolution and adaptation relies on identifying genetic mechanisms of natural selection.
• Existing methods like integrated haplotype score (iHS) and Fay and Wu's H statistic aid candidate gene discovery but struggle to pinpoint causal variants.
• Limitations in localizing causal variants hinder detailed functional studies of positive selection.

Purpose of the Study:

• To develop a novel computational method for detecting recent positive selection.
• To improve the localization accuracy of causal variants within selected genomic regions.
• To provide a computationally efficient tool applicable to large-scale sequencing data.

Main Methods:

• Developed a new method based on conditional coalescent trees to detect positive selection.
• Utilized unbalanced mutation counts on coalescent gene genealogies.
• Validated the method through extensive simulations and analysis of empirical data.

Main Results:

• The novel method demonstrates robustness against demographic biases (bottleneck, expansion, stratification).
• Achieved 20-40% higher accuracy in localizing causal variants compared to state-of-the-art methods on simulated data.
• Successfully identified causal variants in known positively selected genes (ADH1B, MCM6, APOL1, HBB) in empirical data.
• Exhibits significantly higher computational efficiency (24-66x faster than REHH, >10,000x faster than iHS).

Conclusions:

• The conditional coalescent tree method offers a powerful and accurate approach for detecting recent positive selection.
• Superior variant localization capabilities facilitate the identification of functional genetic drivers of human adaptation.
• High computational efficiency makes the method suitable for analyzing large genomic datasets, advancing evolutionary genetics research.