A stochastic model for immune response with mutations and evolution (2404.17950v2)

Abstract: We consider two simple stochastic models for a pathogen population in the presence of an immune response, where we assume that the immune system must first get rid of the pathogen type with the lowest fitness in each ancestral lineage of pathogen types present in the host. In the first model, the pathogens reproduce independently at rate $\lambda$ without spatial restrictions. The second model is similar but evolves on a graph. In this case, each pathogen occupies a site and can only place offspring on adjacent empty sites. In both models, when a new pathogen is born, it has the same type as its parent with probability $1-r$. Conversely, with probability $r$, a beneficial mutation occurs, generating a new pathogen type with a higher fitness than its parent's type. We assume that the immune system develops an effective immune response, independently at rate 1, for each pathogen type without ancestral types that are still alive, which simultaneously kills all pathogens of that type. Our main results are a complete characterization of the survival-extinction phase diagram of the first model, and sufficient conditions for survival and extinction of the second model on $\mathbb{Z}^d.$ A comparison with similar models introduced by Schinazi and Schweinsberg (2008), where mutations are not necessarily beneficial and the immune response for each pathogen type occurs independently at rate 1, reveals that the first and second models exhibit more complex and comparable behaviors, respectively.