Modelling the Effects of Antibiotics on Gut Flora Using a Nonlinear Compartment Model with Uncertain Parameters

We present a mathematical model of a microbial community in the human colon, including transport, exchange and metabolic processes. The colon is represented as a bioreactor with separate lumen and mucus microhabitats. The microbial community in the colon is grouped into four biomass functional groups based on their metabolic activity. While computational models present a challenge in the selection of model parameters, they also provide a more systematic approach to addressing uncertainties when compared to in vivo and in vitro experiments. We conduct an exploratory study on the uncertainty of our input parameters and a simulation study of the perturbation and recovery of gut microbiota from antibiotic exposure. We consider our parameters as random variables drawn from a uniform random distribution to reflect the diversity of gut microbial composition and variability between individuals.