TWAS-GKF: a novel method for causal gene identification in transcriptome-wide association studies with knockoff...

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Transcriptome-wide association studies (TWAS) identify genes linked to traits and their biological mechanisms.
• Current TWAS methods struggle with finite-sample false discovery rate (FDR) control and require individual-level genetic data, which is often unavailable.

Purpose of the Study:

• To develop a novel TWAS method that guarantees finite-sample FDR control.
• To enable gene-trait association discovery using only summary statistics, overcoming data accessibility limitations.

Main Methods:

• Proposed TWAS-GKF, a knockoff inference method utilizing Ghostknockoff principles.
• Generated knockoff variables solely from summary statistics, eliminating the need for individual-level data.
• Validated the method's FDR control and performance in simulations and real-world applications.

Main Results:

• TWAS-GKF demonstrated robust finite-sample FDR control across all tested scenarios.
• Identified significant trait-associated genes in brain cerebellum (schizophrenia) and liver (LDL-C) tissues.
• A majority of identified genes were validated using the Open Targets Validation Platform.

Conclusions:

• TWAS-GKF offers a powerful and accessible approach for identifying trait-associated genes.
• The method effectively controls FDR and leverages summary statistics for broader applicability in genetic association studies.