Correcting population stratification in genetic association studies using a phylogenetic approach

Area of Science:

• Genetics and Genomics
• Population Genetics
• Bioinformatics

Background:

• Genetic association studies are crucial for mapping complex diseases using genotyping data and single nucleotide polymorphisms (SNPs).
• Population-based designs with unrelated individuals are practical but susceptible to errors from unrecognized population stratification.
• Population stratification can lead to false findings and obscure true genetic associations if not properly addressed.

Purpose of the Study:

• To introduce PHYLOSTRAT, a novel computational method for correcting population stratification in genetic association studies.
• To develop a hybrid approach that effectively captures both discrete and admixed population structures.

Main Methods:

• PHYLOSTRAT combines phylogenetic tree construction from SNP genotypes with principal coordinates from multi-dimensional scaling (MDS) analysis.
• This hybrid strategy integrates the hierarchical nature of phylogeny with the admixture-representing capabilities of MDS.

Main Results:

• Simulations and analyses of real datasets demonstrate PHYLOSTRAT's superior efficiency in correcting population stratification compared to existing methods like EIGENSTRAT and STRATSCORE.
• PHYLOSTRAT requires fewer random SNPs for accurate inference of population structure.
• The method effectively captures complex human population structures, including discrete and admixed groups.

Conclusions:

• PHYLOSTRAT offers a robust and efficient solution for correcting population stratification in genetic association studies.
• The method enhances the reliability of gene mapping for complex diseases by mitigating biases from population structure.
• PHYLOSTRAT's hybrid approach provides a flexible and effective way to analyze diverse human population data.