repfdr

repfdr implements an empirical Bayes method to identify single nucleotide polymorphisms (SNPs) that replicate across multiple genome-wide association studies (GWAS).

Key Features:

• Empirical Bayes Approach: Applies an empirical Bayes methodology to detect replicated associations and control false discoveries across studies.
• Flexible Implementation: Supports analysis when multiple investigations test identical sets of null hypotheses, allowing adaptation to varied study designs.
• Meta-Analysis Capabilities: Enables integration and synthesis of results from multiple studies to assess consistency of genetic associations.

Scientific Applications:

• Replicability Analysis in GWAS: Identifying SNPs consistently associated with phenotypes across multiple GWAS to distinguish true from spurious associations.
• Meta-Analysis of Genetic Associations: Combining evidence across studies to increase power and evaluate consistency of SNP–phenotype signals.

Methodology:

Implemented in R, repfdr uses an empirical Bayes statistical framework that pools information across studies to improve SNP effect estimates and identify replicated associations by incorporating prior information across studies.