A0762
Title: Flexible aggregation of microbiome predictors
Authors: Christine Peterson - The University of Texas MD Anderson Cancer Center (United States) [presenting]
Satabdi Saha - University of Texas MD Anderson Cancer Center (United States)
Abstract: Microbiome data sets, which capture the abundances of bacteria and other microorganisms in the human body, represent a key source of "big data" in understanding human health. The purpose is to first introduce the structure of microbiome data and unique challenges in the analysis of this high-dimensional data type. A Bayesian approach is then discussed for feature selection in regression with microbiome predictors. In the proposed model, flexible aggregation of microbiome features is enabled through data-adaptive clustering of their coefficients, allowing for the identification of microbial taxa with similar functional impacts. This method is illustrated with an application to a data set on the role of the microbiome in insulin resistance.
