A General Framework for Powerful Confounder Adjustment in Omics Association Studies

Genomic data are subject to various sources of confounding, such as batch effects and cell mixtures. To identify genomic features associated with a variable of interest in the presence of confounders, the traditional approach involves fitting a confounder-adjusted regression model to each genomic feature, followed by multiplicity correction. Previously, we showed that this procedure was sub-optimal and proposed a more powerful procedure named the two-dimensional false discovery rate control (2dFDR) procedure, which relied on the test statistics from both confounder-adjusted and unadjusted linear regression models (Yi et al (2021)). Though 2dFDR provides significant power improvement over the traditional method, it is based on the linear model assumption that may be too restrictive for some practical settings. This study proposes a model-free two-dimensional false discovery rate control procedure (MF-2dFDR) to significantly broaden the scope and applicability of 2dFDR. MF-2dFDR uses marginal independence test statistics as auxiliary information to filter out less promising features, and FDR control is performed based on conditional independence test statistics in the remaining features. MF-2dFDR provides (asymptotically) valid inference from samples in settings in which the conditional distribution of the genomic variables given the covariate of interest and the confounders is arbitrary and completely unknown. To achieve this goal, our method requires the conditional distribution of the covariate given the confounders to be known or can be estimated from the data. We develop a conditional randomization procedure to simultaneously select the two cutoff values for the marginal and conditional independence test statistics. Promising finite sample performance is demonstrated via extensive simulations and real data applications.