Statistical tests of genetic association for case-control study designs

Area of Science:

• Genetics
• Statistical genetics
• Population genetics

Background:

• Case-control genetic association studies aim to identify genetic markers linked to trait status.
• Existing methods like Pearson's chi-square test and model-based tests have limitations in power and robustness to model misspecification.
• Research focuses on improving the robustness of model-based tests for genetic association.

Purpose of the Study:

• To propose a novel statistical analysis framework for case-control genetic association studies.
• To develop a method that enhances statistical power and robustness against model misspecification.
• To introduce a flexible framework that does not rely on specific trait models or Hardy-Weinberg equilibrium (HWE) assumptions.

Main Methods:

• Developed an analysis framework to test allele frequency equality, allowing for different deviations from Hardy-Weinberg equilibrium (HWE) between cases and controls.
• Introduced likelihood ratio, score, and Wald statistics within this framework.
• The proposed method operates with a single degree of freedom, simplifying analysis.

Main Results:

• The proposed method demonstrated robustness and efficiency in identifying trait-associated genetic markers.
• Computer simulations showed competitive or superior performance compared to existing methods.
• The framework's flexibility allows for analysis without prespecified trait models or HWE conformity.

Conclusions:

• The novel statistical framework offers a powerful and robust approach for genetic association studies.
• It overcomes limitations of existing methods by not requiring trait model specification or HWE assumptions.
• This method provides a valuable tool for identifying genetic markers associated with trait status in diverse populations.