CMStatistics 2020: Start Registration
View Submission - CMStatistics
Title: Compositional mediation models: Application to microbiome data Authors:  Michael Sohn - University of Rochester (United States) [presenting]
Abstract: The importance of the microbiome in maintaining human health and its contribution to disease when it is perturbed has been well established. It is also well known that the microbiome is shaped by external factors, such as diet and medication. Therefore, understanding the mediating role of the microbiome in linking external factors and our health conditions is crucial to translate the microbiome research into therapeutic and preventative applications. We introduce sparse compositional mediation models under potential outcomes framework to estimate and test the causal mediation effect (i.e., mediation effects of the microbiome) utilizing the compositional algebra defined in the simplex space and a linear zero-sum constraint on regression parameters.