Black Box FDR

TL;DR

Black Box FDR is a novel empirical-Bayes method that uses deep learning models to improve the detection of significant results and relevant features in large-scale multi-experiment studies, enhancing power and FDR control.

Contribution

It introduces a two-stage black box modeling approach for multi-experiment analysis, leveraging neural networks to boost discovery power and interpretability.

Findings

Outperforms state-of-the-art methods in benchmarks.

Increases discovery of significant outcomes in cancer studies.

Identifies key genomic drivers of drug response.

Abstract

Analyzing large-scale, multi-experiment studies requires scientists to test each experimental outcome for statistical significance and then assess the results as a whole. We present Black Box FDR (BB-FDR), an empirical-Bayes method for analyzing multi-experiment studies when many covariates are gathered per experiment. BB-FDR learns a series of black box predictive models to boost power and control the false discovery rate (FDR) at two stages of study analysis. In Stage 1, it uses a deep neural network prior to report which experiments yielded significant outcomes. In Stage 2, a separate black box model of each covariate is used to select features that have significant predictive power across all experiments. In benchmarks, BB-FDR outperforms competing state-of-the-art methods in both stages of analysis. We apply BB-FDR to two real studies on cancer drug efficacy. For both studies, BB-FDR increases the proportion of significant outcomes discovered and selects variables that reveal key genomic drivers of drug sensitivity and resistance in cancer.