PowerGxE

PowerGxE performs power calculations as a SAS macro for genetic association tests that jointly assess marginal genetic effects and gene-environment interactions in case-control studies of complex diseases.

Key Features:

• Joint test of marginal association and gene-environment interaction: Implements a joint test that simultaneously evaluates marginal genetic effects and gene-environment interactions in case-control data.
• Scalability to large marker scans: Supports analyses across marker scans ranging from 1,000 to 1,000,000 markers.
• Comparative evaluation of analytical methods: Compares the joint test to marginal genetic association tests, logistic regression-based gene-environment interaction tests, and case-only interaction tests that assume gene-environment independence.
• Power and sample size estimation: Evaluates power and sample size requirements relative to alternative analytical methods.
• Robustness across penetrance models: Demonstrates near-optimal power across a wide range of penetrance models when the true interaction model is unknown.
• Enhanced detection in exposure-specific scenarios: Shows superior power to marginal tests when genetic effects are confined to exposed subjects and superior power to standard GxE tests when genetic effects are not exposure-restricted.

Scientific Applications:

• Large-scale association scans for complex disease loci: Applied to identify susceptibility loci in genome-wide or large-marker scans by integrating marginal and interaction analyses.
• Investigation of gene-environment interplay in multifactorial diseases: Used to assess how genetic variants and environmental exposures jointly influence disease etiology.
• Study design and power/sample-size planning for GxE studies: Used to inform required sample sizes and expected power under varying penetrance and interaction scenarios.

Methodology:

Implemented as a SAS macro that conducts a joint test of marginal association and gene-environment interaction, compares power and sample size to marginal tests, logistic regression-based GxE tests, and case-only tests, and evaluates performance across penetrance models.