Improved score statistics for meta-analysis in single-variant and gene-level association studies

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Meta-analysis is crucial for genetic association studies, accelerating genetic discovery.
• Standard meta-analysis methods face power loss in unbalanced case-control ratio studies.
• Joint analysis is computationally intensive but accurate.

Purpose of the Study:

• To investigate power loss in standard meta-analysis for unbalanced genetic studies.
• To propose novel meta-analysis methods that match joint analysis performance.
• To enhance power and accuracy in genetic association studies.

Main Methods:

• Derived improved meta-score-statistics for linear and logistic regression models.
• Developed a novel approach to adjust for population stratification using minor allele frequencies.
• Simulated gene-level association studies under unbalanced settings.

Main Results:

• Proposed methods recovered up to 85% of power lost by standard methods in simulations.
• Demonstrated power gains in gene-level tests using real-world data (age-related macular degeneration).
• Addressed challenges in meta-analyzing multi-ethnic samples (type 2 diabetes).

Conclusions:

• Improved meta-score-statistics offer a powerful and convenient framework for genetic association studies.
• The novel methods provide accurate approximations to joint analysis, even with unbalanced data.
• Corrections for population stratification enhance the reliability of cross-study analyses.