Learning False Discovery Rate Control via Model-Based Neural Networks

TL;DR

This paper presents a neural network-based method to improve false discovery rate control in high-dimensional variable selection, achieving tighter error control and higher discovery power.

Contribution

It introduces a model-based neural network approach that replaces analytical FDP estimators, enabling more accurate FDR control and increased detection power.

Findings

Achieves tighter FDR control close to the target level.

Demonstrates superior variable detection in simulations.

Effective in complex synthetic GWAS data.

Abstract

Controlling the false discovery rate (FDR) in high-dimensional variable selection requires balancing rigorous error control with statistical power. Existing methods with provable guarantees are often overly conservative, creating a persistent gap between the realized false discovery proportion (FDP) and the target FDR level. We introduce a learning-augmented enhancement of the T-Rex Selector framework that narrows this gap. Our approach replaces the analytical FDP estimator with a neural network trained solely on diverse synthetic datasets, enabling a substantially tighter and more accurate approximation of the FDP. This refinement allows the procedure to operate much closer to the desired FDR level, thereby increasing discovery power while maintaining effective approximate control. Through extensive simulations and a challenging synthetic genome-wide association study (GWAS), we demonstrate that our method achieves superior detection of true variables compared to existing approaches.