Tackling the Excessive Zeros and Phylogenetic Constraint for Microbiome Data

• Activity Number: 445 - Quantification, Association Testing, and Integration of Micriobiome
• Type: Invited
• Date/Time: Wednesday, August 2, 2017 : 8:30 AM to 10:20 AM
• Sponsor: WNAR
• Author(s): Jun Chen*
• Companies: Mayo Clinic

Abstract

Next generation sequencing technologies have enabled the study of the microbiome in a culture-independent way using direct DNA sequencing. One strategy sequences the bacterial 16S rRNA gene for studying the bacterial component of the microbiome. The sequenced 16S tags are usually clustered into the Operational Taxonomic Units (OTUs) based on sequence similarity and downstream statistical analyses are performed on the OTU data. Two major characteristics of OTU data are the excessive zeros due to a large number of rare OTUs and the phylogenetic tree that relates all the OTUs. In this presentation, I will introduce statistical models accounting for these unique characteristics under two important topics: differential abundance analysis and mediation analysis. For differential abundance analysis, I will introduce a generalized regression framework for zeroinflated counts and a prior structure-adjusted false discovery rate control procedure to accommodate the phylogenetic structure. For mediation analysis, I will introduce a nonparametric, distance-based approach, where distance measures are used to address the excessive zeros and tree structure of microbiome data. I will illustrate these methods using simulations as well as real data sets.