Using Genetic Algorithms to Model Microbiome Coevolutionary Dynamics and Dysbiosis Due to Environmental and Pharmaceutical Stressors

Abstract

Several studies have established the critical role of microbiome in shaping human health. Steady state balance, a microbial homeostasis, of disparate microbial colonies is the outcome of coevolution and affects the continued health, chronic disease or susceptibility to ill-health. Environmental stressors, including infection and pharmaceuticals, can trigger imbalance and maladaptation of these microbial colonies. Microbial populations of related species are often associated with a specific biological outcome due to a shared biological function (clustered signal). Similarly, diverse interdependent species are also associated with a specific biological outcome (dense signal). When either deliberate or inadvertent influences disrupt the stable relative population of microbes, understanding the dynamics of coevolution in the altered state is important if we are to ultimately understand the longer-term effects of such a disruption. This study attempts to create a generalized approach to model the coevolutionary dynamics of the microbiome due externally triggered disruptions. Preliminary results suggest that the model is successful in simulating stable relative compositions and evaluating pair-wise competition/cooperation scores for microbiome species. The results support the prospect of simulating and predicting the prevalence of Inflammatory Bowel Disease (IBD) as a result of co-evolutionary dynamics. The results further support the possibility of using such a computational approach to model antibiotic induced disruptions to the microbiome.