A consistent approach to the genotype encoding problem in a genome-wide association study of continuous phenotypes

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) analyze genetic variants for trait associations.
• Population stratification correction can alter genotype data, impacting conventional association tests.
• Pearson's correlation coefficient-based tests are sensitive to genotype encoding methods.

Purpose of the Study:

• To develop a hypothesis testing method robust to different genotype encodings in GWAS.
• To address the dependency of conventional tests on genotype representation after population stratification correction.
• To propose a non-parametric test that is less sensitive to genotype encoding variations.

Main Methods:

• Proposed a non-parametric hypothesis test utilizing Kendall's tau.
• Compared the robustness of Kendall's tau test against Pearson's correlation coefficient test.
• Evaluated test performance using simulated and real genetic datasets with varying genotype encodings.

Main Results:

• Kendall's tau test showed increased robustness across different genotype encodings compared to Pearson's test.
• The p-values from Kendall's tau test exhibited less variability with altered genotype encodings.
• Demonstrated superior performance of Kendall's tau in both simulated and real data analyses.

Conclusions:

• Kendall's tau test offers a more reliable approach for GWAS, particularly when dealing with adjusted genotype data.
• The findings advocate for careful consideration of genotype encoding in numerical analyses within population and comparative genomics.
• The proposed method enhances the association analysis of novel genetic variants with traits.