An evolution model with uncountably many alleles

We study a class of evolution models, where the breeding process involves an arbitrary exchangeable process, allowing for mutations to appear. The population size n is fixed, hence after breeding, selection is applied. Individuals are characterized by their genome, picked inside a set X (which may be uncountable), and there is a fitness associated to each genome. Being less fit implies a higher chance of being discarded in the selection process. The stationary distribution of the process can be described and studied. We are interested in the asymptotic behavior of this stationary distribution as n goes to infinity. Choosing a parameter lambda > 0 to tune the scaling of the fitness when n grows, we prove limiting theorems both for the case when the breeding process does not depend on n, and for the case when it is given by a Dirichlet process prior. In both cases, the limit exhibits phase transitions depending on the parameter lambda.