R2-Based Mediation Analysis with High-Dimensional Omics Mediators

• Activity Number: 230 - Recent Advances in Statistical Methods for Omics Data
• Type: Topic Contributed
• Date/Time: Tuesday, August 9, 2022 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Peng N/A Wei* and Tianzhong Yang and Sunyi N/A Chi and Zhichao Xu and Chunlin Li and Bin Shi and Xuelin Huang
• Companies: The University of Texas MD Anderson Cancer Center and University of Minnesota and The University of Texas MD Anderson Cancer Center and The University of Texas MD Anderson Cancer Center and University of Minnesota and The University of Texas MD Anderson Cancer Center and The University of Texas MD Anderson Cancer Center
• Keywords: High-dimensional mediators; mediation analysis; R2; survival analysis; gene expression

Abstract

Environmental exposures can regulate intermediate molecular phenotypes, such as the transcriptome, metabolome and microbiome, by various mechanisms and thereby lead to different health outcomes. It is of significant scientific interest to unravel the role of potentially high-dimensional intermediate phenotypes in the relationship between environmental exposure and health traits. Mediation analysis is an important tool for investigating such relationships. However, there are many unique challenges facing high-dimensional mediation analysis of these emerging “omics” mediators. To this end, we extended an R-squared (R2) total mediation effect size measure for continuous outcomes, originally proposed in the single-mediator setting, to the moderate- and high-dimensional mediator settings in the mixed model framework (Yang et al, BMC Bioinformatics 2021). I will introduce some recent advances in R2-based mediation analysis with high-dimensional omics mediators, including speeding up confidence interval estimation with an asymptotic formula, extension to time-to-event outcomes, and real data applications to the Framingham Heart Study.