Online Program Home
My Program

Abstract Details

Activity Number: 403 - SPAAC Poster Competition
Type: Topic Contributed
Date/Time: Tuesday, July 30, 2019 : 2:00 PM to 3:50 PM
Sponsor: Section on Statistics in Genomics and Genetics
Abstract #304945
Title: Integrative Modeling of Multi-Omic Data Using a Mediation Framework
Author(s): Ilana Trumble* and Daniel Frank and Vijay Ramakrishnan and Miranda Kroehl
Companies: University of Colorado Denver and University of Colorado Anschutz Medical Campus, Department of Medicine and University of Colorado Anschutz Medical Campus, Department of Otolaryngology and Colorado School of Public Health
Keywords: Causal mediation; High-dimensional statistics; Multiple mediation; Integrative genomics

Causal mediation analysis is an increasingly popular method for describing the mechanism through which an independent variable affects an outcome. This framework has been extended to accommodate high-dimensional exposures or mediators. However, methods are limited since they do not allow for a comprehensive analysis of multiple predictors and mediators concurrently. We present a novel modeling approach for integrating high-dimensional exposures and mediators by extension of prior methods under the causal mediation framework, with application to microbiome and metabolomic data. We propose an overall global test for evaluating mediation and employ LASSO regression for parameter estimation and significance testing. Additionally, we propose and evaluate taxon-specific estimators for the indirect effect (IE) of the microbiome–metabolome–outcome relationship. Both tests are evaluated for type I error and power under simulation studies. These methods are applied to an existing data set of chronic rhinosinusitis (CRS) patients to identify and quantify microbiome–metabolome pathways that potentially mediate the relationship between the microbiome and CRS severity.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program