Joint Local False Discovery Rate in GWAS

Posted on Nov 12, 2022 (Update: Nov 14, 2022)

Tags: False Discovery Rate, Genome-wide Association Studies

This note is for Jiang, W., & Yu, W. (2017). Controlling the joint local false discovery rate is more powerful than meta-analysis methods in joint analysis of summary statistics from multiple genome-wide association studies. Bioinformatics, 33(4), 500–507.

\[\newcommand\Jlfdr{\mathrm{Jlfdr}} \newcommand\Fdr{\mathrm{Fdr}}\]

propose a novel summary-statistics-based joint analysis method based on controlling the joint local false discovery rate (Jlfdr).

• prove that the method is the most powerful summary-statistics-based joint analysis method when controlling the false discovery rate at a certain level
• the Jlfdr-based method achieves higher power than commonly used meta-analysis methods when analyzing heterogeneous datasets from multiple GWASs

two kinds of joint analysis methods:

• individual-level
• summary-statistics-based

Jlfdr generalizes the concept of the local false discovery rate (Jlfdr) from the analysis of single study to the joint analysis of multiple studies

Methods

Jlfdr and optimal rejection region

\[\Jlfdr(z) = P(H_0\mid \bfz)\]

Fdr is the expectation of Jlfdr, given that the test statistic vector is in the rejection region $R$.

\[\Fdr(R) = E(\Jlfdr(z)\mid z\in R)\]

Implementation of Jlfdr-based method under the Gaussian mixture model