A powerful procedure that controls the false discovery rate with directional information

2025-07-21

AI Summary

In genetic multiple testing, conventional false discovery rate (FDR) control methods rely solely on p-values and ignore the biological directionality encoded in the signs of test statistics (e.g., gene upregulation vs. downregulation), resulting in suboptimal statistical power. To address this, we propose *signed-knockoffs*, the first finite-sample knockoff framework that explicitly incorporates directional information. Our method constructs signed, direction-aware knockoff variables and designs asymmetric rejection regions to enable direction-sensitive FDR control. It requires no additional distributional assumptions and provides rigorous theoretical guarantees alongside computational feasibility. Extensive simulations and analyses of real gene expression datasets demonstrate that signed-knockoffs maintain the target FDR level while substantially improving detection power for both up- and down-regulated genes—outperforming classical p-value–based approaches.

Technology Category

• Machine Learning: Kernel Methods
• Search and Optimization: Mixed Discrete/Continuous Search
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In many multiple testing applications in genetics, the signs of test statistics provide useful directional information, such as whether genes are potentially up- or down-regulated between two experimental conditions. However, most existing procedures that control the false discovery rate (FDR) are $p$-value based and ignore such directional information. We introduce a novel procedure, the signed-knockoff procedure, to utilize the directional information and control the FDR in finite samples. We demonstrate the power advantage of our procedure through simulation studies and two real applications.

Problem

• Controls FDR with directional information in genetics
• Addresses limitations of p-value based FDR methods
• Proposes signed-knockoff procedure for finite sample FDR control

Innovation

• Controls FDR with directional information
• Uses signed-knockoff procedure
• Validates power via simulations