Estimation of Interaction and Growth Parameters to Develop a Computational Model for Gut Bacteria

Abstract

AbstractThe relevance of gut bacterial balance to human health can not be overemphasized. The gut bacterial balance delicately relies on several factors inherent to the person as well as to the environment. As the volume of evidences for the gut bacterial influence on health and the clinical data on the variance of the bacterial population across cohorts continue to grow exponentially, it is important to develop a theoretical model for gut bacteria. In this work, we suggest a new computational method for estimating the interaction parameters from the cross-sectional data of bacterial abundances in a cohort, without requiring a longitudinal followup. We introduce a nutrient type based bacterial growth model and use the Monte Carlo approach to estimate the matrix of interaction parameters for the 14 major bacterial species. These parameters were used in a comprehensive first-level computational model we developed for the large intestine to understand the patterns of re-establishing balance with different nutrient types.