Our goal is to use the intestinal microbiota as a model and apply genetic and environmental perturbations to generate variations in gut microbiota (GM) composition and then use integrative genomics approaches to infer interactions among host and microbiota phenotypes. To reach our goal, we are developing enabling approaches in several critical areas: 1. Determine the best combinations of perturbation to use. We are considering both genetic and environmental perturbations. While existing data point to the importance of both environmental and genetic factors in shaping the microbiota, it is unclear, especially given the myriad confounding factors (e.g., cage and breeding environments), whether one is significantly more dominant than the other and what combined effects they exert together. Using mouse as a model, we are in the process of evaluating this and design appropriate perturbation and breeding strategies. 2. Develop inexpensive and high-throughput strategies to assay microbiome phenotypes and data analysis methods to process and analyze data generated from the individual data types. Phenotypes of interest include microbial composition and DNA and transcript contents. A first goal is to develop an Illumina-based approach to profile microbiota composition of complex samples (e.g., those from the gut). Existing culture-free techniques primarily use 454 pyrosequencing, which typically provides 10-20k reads per sample but still at a considerable cost per sample. Using the Illumina Hiseq platform would allow us to achieve significant deeper coverage (500k reads per sample) with lower costs ($20-30/sample). We have successfully developed a PCR amplicon based approach to sequence multiple 16s and genomics regions from samples. By using a barcoding and custom-primer strategy, we have conducted massively parallel phenotyping of microbiota composition of a number of intestinal, fecal and skin samples (in collaboration with Yasmine Belkaid). Some of these have already been sequenced by 454 to allow comparison. In collaboration with colleagues at the Computational Biology Branch of the Office of Cyber Infrastructure and Computational Biology, we are developing and testing Illumina-based data analysis methods for inferring microbial diversity and relative abundance by integrating information across genomic regions.
