Methods for modeling the dynamics of microbial communities

Advances in DNA sequencing of microbial communities have revealed a complex relationship between the human microbiome and our health. Community dynamics, host-microbe interactions, and changing environmental pressures create a dynamic ecosystem that is just beginning to be understood. In this work, we develop methods for investigating the dynamics of the microbiome. First, we develop a model for describing community dynamics. We show that the proposed approaches accurately describes community trajectories over time. Next, we develop a method for modeling and eliminating technical noise from longitudinal data. We demonstrate that the method can accurately reconstruct microbial trajectories from noisy data. Finally, we develop a method for estimating bacterial growth rates from metagenomic sequencing. Using a case-control cohort of individuals with irritable bowel disease, we show how growth rates can be associated with disease status, community states, and metabolites. Altogether, these models can be used to help uncover the relationship between microbial dynamics, human health, and disease.