Coevolution in action in host and parasite genomes

点击翻译按钮获取中文摘要

When hosts and pathogens exert strong reciprocal selection pressure, the genes underpinning the interaction coevolve over short time scales, while populations undergo steep year-to-year size fluctuations. Surprisingly, however, no method was ever developed to track genes undergoing coevolutionary dynamics using serial sampling of local host and pathogen populations over time. We have developed a first Approximate Bayesian Computation (ABC) method which allows to reveal the genetic underpinnings of coevolutionary dynamics using jointly host and parasite whole-genome polymorphism data. The coevolutionary dynamics of important genes for the interaction can be distinguished from demographic dynamics of host and pathogen populations by assessing and modelling genetic diversity over time and over the whole genome. The method is thus applicable to serial samplings of one given coevolving host and pathogen population.The goal of this project is to extend this method to study more complex but also realistic set ups of coevolution: 1) host and parasite spatial structure with and without heterogeneity in selection, and 2) violation of the classic Wright-Fisher model assumption in parasite species. First, we study coevolution in space by using host and parasite genome polymorphism data from several populations. We will build a spatial homogeneous and heterogeneous host and parasite population model, in which migration and local changes in population size can be inferred using full genome data. The ABC method will then be used to search for genes under coevolution and infer the parameters of the geographic mosaic of coevolution. It is of special interest to assess if sampling of several populations in space is equivalent to time sampling used in our current method.Second, parasite due to their life cycles may violate the assumptions of our models in two ways: 1) parasites show large variation in their population size over time generating recurrent bottlenecks, and 2) parasite species can exhibit large variance in offspring distribution. Our inference method may be biased when dealing with parasites such as viruses or bacteria. We will therefore integrate these two key aspects of parasite life-cycles into our model and our ABC inference method, and study how they affect the coevolutionary dynamics and polymorphism at the coevolving loci and genome-wide. We therefore aim to provide a novel inference method applicable to most host-parasite systems for which full genome data are available for several populations and several time points.