Your browser is unsupported

We recommend using the latest version of IE11, Edge, Chrome, Firefox or Safari.

Sep 26 2022

9/26 – Yinglin Xia, Department of Medicine, UIC

CBQB Seminar (Virtual Event)

September 26, 2022

12:00 PM - 1:30 PM


CBQB Seminar (Virtual Event)

This is a virtual event you can also watch seminar live here>>

Yinglin Xia, PhD
Research Professor
Department of Medicine, University of Illinois at Chicago

Microbiome data characteristics and statistical analysis models

“Microbiome” is defined as the microbial taxa or microbes and their genes--the whole microbial communities and their activities. Microbiome data are typically generated by two approaches: either 16S rRNA and shotgun metagenomics. By either approach, the generated data have some unique data structure and characteristics, such as microbiome data are classified into hierarchical taxonomic ranks and encoded as a phylogenetic tree, multivariate or high dimensional, compositional, over-dispersed, often sparse with many zeros, and heterogeneous. All those unique characteristics pose major challenges for statistical analysis and modeling of microbiome data using standard statistical methods: 1) high-dimensionality causes the large P and small N problem; 2) compositionality causes the dependency problem; 3) sparsity with excess zeros causes the over-dispersion and zero-inflation problems; and 4) heterogeneity challenges data integration, modeling, and meta-analysis.

To target these unique data structure and characteristics, and to address the challenges, many statistical models/methods have been adopted from classical statistical field and other related fields as well as developed in current years. In this talk, we will describe the microbiome research themes and review both cross-sectional and longitudinal statistical models/methods for microbiome data. A non-parametric multivariate analysis of microbiome data using PERMANOVA will be illustrated.



UIC Biomedical Engineering

Date posted

Sep 12, 2022

Date updated

Sep 12, 2022