Modeling Between-Study Heterogeneity for Improved Replicability in Gene Signature Selection and Clinical Prediction

Abstract

In the genomic era, the identification of gene signatures associated with disease is of significant interest. Such signatures are often used to predict clinical outcomes in new patients and aid clinical decision-making. However, recent studies have shown that gene signatures are often not replicable. This occurrence has practical implications regarding the generalizability and clinical applicability of such signatures. To improve replicability, we introduce a novel approach to select gene signatures from multiple datasets whose effects are consistently non-zero and account for betweenstudy heterogeneity. We build our model upon some rank-based quantities, facilitating integration over different genomic datasets. A high dimensional penalized Generalized Linear Mixed Model (pGLMM) is used to select gene signatures and address data heterogeneity. We compare our method to some commonly used strategies that select gene signatures ignoring between-study heterogeneity. We provide asymptotic results justifying the performance of our method and demonstrate its advantage in the presence of heterogeneity through thorough simulation studies. Lastly, we motivate our method through a case study subtyping pancreatic cancer patients from four gene expression studies.

Hierachical Resampling for Bagging in Multi-Study Prediction with Applications to Human Neurochemical Sensing

“…Here we propose hierarchical resampling techniques coupled with covariate distributionbased weighting schemes towards similar ends. We leverage ideas from the rich literature in transfer learning and the rapidly growing "multi-study" statistics perspective that proposes methods to combine studies in supervised (Guan, Parmigiani and Patil, 2020;Ramchandran, Patil and Parmigiani, 2020;Ren et al, 2021), unsupervised (De Vito et al, 2019;Roy et al, 2019) and inference settings (Guo et al, 2021;Rashid et al, 2020).…”

Linear Hypothesis Testing in Ultra High Dimensional Generalized Linear Mixed Models

“…Our manuscript is organized as follows. Section 2.2 reviews the pGLMMs modeling framework, first described in Rashid et al (2020). Section 2.3 describes the MCECM algorithm used by glmmPen to fit pGLMM models.…”

Meta-Analyzing Multiple Omics Data With Robust Variable Selection

“…They treated the coefficients of each covariate from all datasets as groups, and performed the simultaneously variable selection both on the group and within the group. For other existing variable selection methods including, for example, group Bridge, composite MCP, and group exponential lasso that can be extended to meta-analyzing multiple studies, one may refer to Zhao et al ( 2015 ), Kim et al ( 2017 ), and Rashid et al ( 2020 ).…”