The human gut microbiome is a tremendously complex ecosystem with a marked impact on our health. The specific mix of species in this ecosystem varies substantially across individuals, over time, and in association with disease. These gut-dwelling species are endowed with diverse metabolic capacities and continuously break down and synthesize numerous metabolites. These capacities, however, and consequently the synthesis of metabolites in the gut, depend on the composition of species in the microbiome and on the web of metabolic interactions between microbiome members. Another important factor that modulates metabolism in the gut is the host diet. Indeed, diet has been shown to play a key role in shaping both the microbiome?s composition and the abundance of various gut metabolites. Yet, the relationship between the composition of the microbiome, the gut metabolites, and the host diet is nontrivial and extremely complex. To address this challenge, numerous studies are now taking a multi-omic approach, harnessing the progress in high-throughput technologies to profile different facets (such as species, gene, and metabolite compositions) of each microbiome sample and to identify associations between these facets. This approach, however, ignores the wealth of knowledge about the mechanisms that link microbiome ecology and metabolism and fails to model such mechanisms, and therefore the findings obtained often lack clear mechanistic interpretations. Our proposed research aims to introduce two novel computational frameworks, integrating multi-omic data with various metabolic modeling approaches, to model the link between microbiome composition, gut metabolites, and diet and to provide a more mechanistic, comprehensive understanding of these relationships. Our first framework will focus on the relationship between the composition of the microbiome and its impact on gut metabolites. It will use taxonomic, genomic, and enzymatic data to model community-wide metabolism and to estimate the community?s potential to synthesize/degrade each metabolite. We will analyze several large-scale datasets pairing microbiome and metabolomic assays to examine how well communities? estimated metabolic potentials explain observed variation in the gut metabolome. We will further develop methods for identifying species that drive this variation and universal mechanisms governing this relationship. Our second framework will focus on the impact of diet and dietary interventions on microbiome composition. This framework will use taxonomic, genomic, and nutritional data to construct models of community members, to convert dietary information to metabolite intake, and to utilize a novel multi-species dynamic metabolic modeling approach, aiming to predict the growth and metabolism of community members over time on a given diet. We will apply this framework to predict diet-induced microbiome compositions using data from several studies that assayed the microbiome response to well-defined diets. We will further use this framework to explore the metabolic mechanisms that underlie this complex diet-microbiome relationship.
Our gut microbiome and our diet both play a key role in our health, yet the complex relationship between the ecology of the microbiome and the composition of the diet, as well as the way they jointly impact metabolic activity in the gut, are poorly understood. In this project, we will develop a suite of novel computational frameworks, integrating several metabolic modeling approaches with metagenomic, metabolomic, and dietary data, to elucidate the mechanisms underlying this complex interplay between gut microbiome, gut metabolites, and diet. We will specifically aim to identify microbial drivers of observed variation in the gut metabolic niche and to predict the impact of dietary shifts on microbiome composition, ultimately obtaining a more mechanistic understanding of this complex system and informing future efforts for microbiome-based therapies and nutritional interventions.
| Verster, Adrian J; Borenstein, Elhanan (2018) Competitive lottery-based assembly of selected clades in the human gut microbiome. Microbiome 6:186 |
| McNally, Colin P; Borenstein, Elhanan (2018) Metabolic model-based analysis of the emergence of bacterial cross-feeding via extensive gene loss. BMC Syst Biol 12:69 |
