Advanced Approaches to Characterize the Human Intestinal Microbiota by Computational Meta-analysis

Abstract

We describe advanced approaches for the computational meta-analysis of a collection of independent studies, including over 1000 phylogenetic array datasets, as a means to characterize the variability of human intestinal microbiota. The human intestinal microbiota is a complex microbial community, consisting of several thousands of phylotypes, is specific for each individual, and impacts health and disease. We have developed a phylogenetic microarray, the Human Intestinal Tract Chip, to address the microbial diversity of the intestinal microbiota and used this tool to generate large datasets. It is of significant interest to use these datasets to be able to provide relations between microbial taxa, describe the extent and type of variability of the microbiota in the human gut, and establish relations between microbial taxa and their interaction with the host, intestinal location, or genotype. We present an advanced computational meta-analysis approach for studying human intestinal microbiota, outline the advantages and disadvantages of such a meta-analysis, and reflect it to analogous approaches in other fields. Finally, we illustrate the potential of this meta-analysis by identifying salient signatures of site-specific microbial communities, describe impact of genotype, and provide first examples of relevant relations between microbial taxa. We are in the process of designing and applying appropriate methods for carrying out a full meta-analysis of the present data. Beyond that, the next large challenges in future meta-analyses lie in the integration of data from several heterogeneous measurement methods such as next generation sequencing techniques, metaproteomics, or metabolomics. We have shown the feasibility of an advanced computational meta-analysis of the large datasets derived from the human intestinal microbiota.