Detecting rare variants in case-parents association studies.

Area of Science:

• Genetics
• Statistical Genetics
• Disease Association Studies

Background:

• Genome-wide association studies (GWASs) primarily detect common variants, but rare variants also influence disease.
• Existing methods struggle with stratification caused by rare variants.
• Case-parent studies offer a robust design for genetic association analysis.

Purpose of the Study:

• To develop and compare methods for testing group-wise associations between multiple rare and common variants in gene regions and disease.
• To assess the performance of these methods in case-parent studies, ensuring robustness against population stratification.

Main Methods:

• Focused on case-parent studies to leverage parent-offspring transmission data.
• Compared four group-wise association tests: single-variant TDT, multivariable test, combined TDT, and a random-effects likelihood ratio test (LRT).
• Utilized extensive simulation studies to evaluate test power and performance under various scenarios.

Main Results:

• The likelihood ratio test demonstrated superior power across a wide range of simulation settings.
• Including common variants alongside rare variants did not negatively impact the LRT's power.
• The LRT remained largely insensitive to the directionality of variant effects when analyzing both deleterious and protective variants simultaneously.

Conclusions:

• The likelihood ratio test is a powerful and robust method for detecting gene-region associations with diseases, particularly in case-parent studies.
• This LRT approach effectively handles population stratification and can incorporate both rare and common variants.
• The LRT offers a reliable tool for dissecting the genetic architecture of diseases influenced by multiple variants.