Dynamic multi-scale modeling of virus-host coevolutionary dynamics
College: Liberal Arts and Sciences
Award year: 2012-2013
Host-pathogen interactions play an important role in the ecology and evolution of all organisms. Eco-evolutionary modeling of the coevolutionary dynamics provides an essential tool to explore the parameters that define these interactions in complex natural systems. For microorganisms, one of the primary diversifying mechanisms is interaction with their lytic viral pathogens. The recently discovered CRISPR adaptive immune system in bacteria and archaea that specifically recognizes DNA of an invading genetic element transforms our understanding of the evolutionary dynamic between microorganisms and their genetic pathogens. In addition, the presence of adaptive immunity makes microbial systems potential models for studying the coevolutionary dynamics of host-pathogen interactions. In order to predict the way that the CRISPR immunity defines host-virus evolutionary dynamics, we have developed a multi-scale model of host-viral evolution that integrates molecular, ecological, and evolutionary parameters. The central conceptual innovation is that the model autonomously changes the dimensionality of a system of coupled ordinary differential equations (ODEs). During large-scale simulations containing tens of thousands of strains over hundreds of replicate populations, genomic states of hosts and viruses are tracked so that the effect of viral immunity on patterns of diversification in both hosts and viruses can be monitored through time. Preliminary simulations demonstrate that multiple coevolutionary dynamics occur, including a new dynamic characterized by invasions of multi-strain coalitions that dramatically preserve diversity and stability in microbial communities. We propose to extend our analysis of the molecular and ecological drivers of these dynamics utilizing supercomputing resources. We will perform large-scale simulations guided by sloppy-stiff parameter searches that allow us to vary parameters in a systematic way as defined by our eco-evolutionary model. We will explore means to extend our simulation approach so as to make it readily available to other researchers using HPC platforms.