An Experimentally-validated Agent-based Model to Study the Emergent Behavior of Bacterial Communities

Abstract

Swimming bacteria are ubiquitous in aqueous environments ranging from oceans to fluidic environments within a living host. Furthermore, engineered bacteria are being increasingly utilized for a host of applications including environmental bioremediation, biosensing, and for the treatment of diseases. Often driven by chemotaxis (i.e. biased migration in response to gradients of chemical effectors) and quorum sensing (i.e. number density dependent regulation of gene expression), bacterial population dynamics and emergent behavior play a key role in regulating their own life and their impact on their immediate environment. Computational models that accurately and robustly describe bacterial population behavior and response to environmental stimuli are crucial to both understanding the dynamics of microbial communities and efficiently utilizing engineered microbes in practice. Many existing computational frameworks are finely-detailed at the cellular level, leading to extended computational time requirements, or are strictly population scale models, which do not permit population heterogeneities or spatiotemporal variability in the environment. To bridge this gap, we have created and experimentally validated a scalable, computationally-efficient, agent-based model of bacterial chemotaxis and quorum sensing (QS) which robustly simulates the stochastic behavior of each cell across a wide range of bacterial populations, ranging from a few to several hundred cells. We quantitatively and accurately capture emergent behavior in both isogenic QS populations and the altered QS response in a mixed QS and quorum quenching (QQ) microbial community. Finally, we show that the model can be used to predictively design synthetic genetic components towards programmed microbial behavior.