Retrospective versus prospective score tests for genetic association with case-control data

Area of Science:

• Genetics
• Biostatistics
• Epidemiology

Background:

• Prospective logistic likelihood is standard for case-control genetic studies.
• Existing aggregated tests for multiple markers often assume common random effects.
• Rare variants and small effects pose challenges in genetic association studies.

Purpose of the Study:

• To develop a novel, powerful aggregation test for genetic association studies.
• To investigate the utility of disease prevalence information for enhancing testing power.
• To address limitations of common-random-effect assumptions in existing methods.

Main Methods:

• Developed a locally most powerful aggregation test using retrospective likelihood.
• Incorporated an independent-random-effect assumption, allowing genetic effects to vary.
• Utilized disease prevalence data to improve statistical power.

Main Results:

• The proposed method demonstrates superior performance compared to existing approaches.
• Simulations confirm the advantages of the new test in various scenarios.
• Disease prevalence information significantly enhances testing power.

Conclusions:

• The new retrospective likelihood-based test offers enhanced power for genetic association studies.
• The independent-random-effect model provides a more flexible framework.
• The method is illustrated with a real genome-wide association study dataset.