METAGEN

METAGEN performs meta-analysis of genome-wide association studies (GWAS) by applying fixed- and random-effects logistic regression models to summary data and treating multiple genotypes as independent variables to enable robust association testing.

Key Features:

• Robust Statistical Methods: Implements Cochran-Armitage Trend Test (CATT) under recessive, additive, and dominant inheritance models, the Maximum Efficiency Robust Test statistic (MAX), and the MIN2 statistic to maximize power while preserving nominal type I error.
• Modeling Framework: Uses fixed- and random-effects logistic regression and explicitly treats multiple genotypes as independent variables for association analysis.
• Asymptotic Distribution Calculation: Calculates asymptotic null distributions for MAX and MIN2 using numerical integration.
• Meta-Analysis Weighting: Supports meta-analysis with summary data using weights based on the reciprocal of combined cases and controls in both fixed and random effects settings.
• Implementation: Implemented in the statistical package Stata.

Scientific Applications:

• GWAS Analysis: Provides robust testing strategies to address unknown genetic models in GWAS and improve detection of genotype–phenotype associations.
• Meta-Analysis: Enables synthesis of summary GWAS data across multiple studies using fixed and random effects to assess genetic associations across populations.

Methodology:

Implements CATT for recessive, additive, and dominant models, MAX and MIN2 statistics; derives asymptotic null distributions for MAX and MIN2 via numerical integration; applies these tests in fixed- and random-effects meta-analysis using summary data weighted by the reciprocal of combined cases and controls; implemented in Stata.