weighted_FDR

weighted_FDR applies a weighted false discovery rate to multiple hypothesis testing to increase statistical power for detecting genetic associations in genome-wide association studies and association genome scans.

Key Features:

• Multiple Testing Problem: Addresses loss of power from stringent corrections such as Bonferroni when hundreds of thousands of polymorphisms are tested.
• False Discovery Rate (FDR) Principle: Leverages FDR control to balance type I error control with higher statistical power compared to traditional methods.
• Weighted Hypotheses: Allows hypotheses to be weighted using prior data, for example from previous linkage scans, to enhance detection of modest genetic effects.
• Linkage Data Utilization: Uses linkage data to assign weights to association P values, integrating prior information into association testing.
• Power Improvement: Demonstrates substantial power gains when linkage studies are informative and minimal power loss when linkage information is limited.
• Sample Size Calculation: Provides a framework to calculate the sample size needed to obtain useful prior information from linkage studies.

Scientific Applications:

• Genome-wide association studies (GWAS): Improves detection of associations between genetic markers and complex diseases by managing the multiple testing burden.
• Association genome scans: Enables exploration of genomic regions with greater confidence even when prior evidence from linkage studies is weak.

Methodology:

Implements a weighted FDR procedure for multiple hypothesis testing that assigns weights to association P values using linkage data and includes a sample size calculation framework for obtaining prior information from linkage studies.