Muller's ratchet model enables to analyse the adaptation stability of a population when deleterious mutations regularly affects new individuals. Here, we condition on the fact that the fittest individuals are still present in the population. Our purpose is to ensure the exponential quasi-ergodicity of the processes (with or without an upper-bound on the number of accumulating deleterious mutations). We also expect to ensure that individuals with too many deleterious mutations play a negligible role both for the survival of the fittest subpopulation and for the quasi-ergodicity of the population on a whole.

We study the long-time behavior of an additive functional that takes into account the jumps of a symmetric Markov process. This process is assumed to be observed through a biased observation scheme that includes the survival to events of extinction and the Feynman-Kac weight by another similar additive functional. Under conditioning for the convergence to a quasi-stationary distribution and for two-sided estimates of the Feynmac-Kac semigroup to be obtained, we shall discuss general assumptions on the symmetric Markov process. For the law of additive functionals, we will prove a quasi-ergodic theorem, namely a conditional version of the ergodic theorem and a conditional functional weak law of large numbers. As an application, we also establish a large deviation principle for the mean ratio of additive functionals.

Our goal is to compare the efficiency of strategies applied on a regional scale to control the spread of a pathogen between urban units, in the following called cities. We model the infection dynamics on a random network where each node represents a city. We take the size of a city as a parameter for its connectivity to other cities. Once a certain threshold number of citizens in a city gets infected the city is isolated to prevent further spread. We consider two different strategies to define the threshold depending on the city size. In strategy $(P)$ a proportion $p$ of citizens need to get infected before isolation, while in strategy $(U)$ an uniform threshold $L$ is chosen for all cities. If city sizes were constant both strategies would coincide. However, in reality city size distributions are heavy-tailed. In this case in strategy $(P)$ the infection of a large city can lead to a much larger number of secondary infections of cities than in strategy $(U)$. To quantify the extend to which these infections fuel the epidemic we benchmark the two strategies against each other by comparing the number of individuals that get isolated when the number of individuals that get infected is under both strategies the same. Furthermore, we define and determine a basic reproduction numbers $R^{(U)}_0$ and $R^{(P)}$ for variant (P) and (U) that represent the initial strengths at which the infection spreads between cities. We fit our model to mobility data from France, Japan and Poland to identify the parameter regimes of our model that are empirically relevant. Furthermore, we investigate by means of simulations to which extend the geographic structure of the city distribution (that we neglect in our model) influences the infection dynamics between cities.

We analyze the long-term stability of a stochastic model designed to illustrate the adaptation of a population to variation in its environment. A piecewise-deterministic process modeling adaptation is coupled to a Feller logistic diffusion modeling population size. As the individual features in the population become further away from the optimal ones, the growth rate declines, making population extinction more likely. Assuming that the environment changes deterministically and steadily in a constant direction, we obtain the existence and uniqueness of the quasi-stationary distribution, the associated survival capacity and the Q-process. Our approach also provides several exponential convergence results (in total variation for the measures). From this synthetic information, we can characterize the efficiency of internal adaptation (i.e. population turnover from mutant invasions). When the latter is lacking, there is still stability, but because of the high level of population extinction. Therefore, any characterization of internal adaptation should be based on specific features of this quasi-ergodic regime rather than the mere existence of the regime itself.

This paper tackles the issue of establishing an upper-bound on the asymptotic ratio of survival probabilities between two different initial conditions, asymptotically in time for a given Markov process with extinction. Such a comparison is a crucial step in recent techniques for proving exponential convergence to a quasi-stationary distribution. We introduce a weak form of the Harnack inequality as the essential ingredient for such a comparison. This property is actually a consequence of the convergence property that we intend to prove. Its complexity appears as the price to pay for the level of flexibility required by our applications, notably for processes with jumps on a multidimensional state-space. We show in our illustrations how simply and efficiently it can be used nonetheless. As illustrations, we consider two continuous-time processes on $\mathbb{R}^d$ that do not satisfy the classical Harnack inequalities, even in a local version. The first one is a piecewise deterministic process while the second is a pure jump process with restrictions on the directions of its jumps.

We establish sufficient conditions for exponential convergence to a unique quasi-stationary distribution in the total variation norm. These conditions also ensure the existence and exponential ergodicity of the Q-process, the process conditioned upon never being absorbed. The technique relies on a coupling procedure that is related to Harris recurrence (for Markov Chains). It applies to general continuous-time and continuous-space Markov processes. The main novelty is that we modulate each coupling step depending both on a final horizon of time (for survival) and on the initial distribution. By this way, we could notably include in the convergence a dependency on the initial condition. As an illustration, we consider a continuous-time birth-death process with catastrophes and a diffusion process describing a (localized) population adapting to its environment.

Starting from a stochastic individual based description of an SIS epidemic spreading on a random network, we study the dynamics when the size of the network tends to infinity. We recover in the limit an infinite-dimensional integro-differential equation studied by Delmas, Dronnier and Zitt (2022) for an SIS epidemic propagating on a graphon. Our work covers the case of dense and sparse graphs, when the number of edges is of order Θ(na) with a>1, but not the case of very sparse graphs with a=1. In order to establish our limit theorem, we have to deal with both the convergence of the graph to the graphon and the convergence of the stochastic process spreading on top of these random structures: in particular, we propose a coupling between the process of interest and an epidemic that spreads on the complete graph but with a modified infectivity.

Comments:
Keywords: random graph, mathematical model for epidemiology, measure-valued process, large network limit, limit theorem, dense graph, sparse graph.
Acknowledgments: This work was financed by the Labex Bézout (ANR-10-LABX-58) and the platform MODCOV19 of the National Institute of Mathematical Sciences and their Interactions of CNRS. 30 pages, including an appendix of 5 pages

Aiming at a simplified model where natural selection at the individual level is confronted with selection effects at the group level, we consider certain individual-based models of large populations subdivided into a large number of groups. We then obtain the convergence to the law of a stochastic process with a certain Feynman-Kac penalization. To analyze the limiting behavior of this law, we exploit a recent approach, designed for the convergence to quasi-stationary distributions. We are able to cope with the absorption of the stochastic process and to relate the convergence to equilibrium to the one where the absorption implies extinction. We notably establish different convergence regimes. Besides the case of an exponential rate (the rate being uniform over the initial condition), critical regimes with convergence in 1/t are also of note. We finally address the relevance of such convergence regimes to predict the long term behavior of the individual-based models.

This article is a presentation of specific recent results describing scaling limits of individual-based models. Thanks to them, we wish to relate the time-scales typical of demographic dynamics and natural selection to the parameters of the individual-based models. Although these results are by no means exhaustive, both on the mathematical and the biological level, they complement each other. Indeed, they provide a viewpoint for many classical time-scales. Namely, they encompass the timescale typical of the life-expectancy of a single individual, the longer one wherein a population can be characterized through its demographic dynamics, and at least four interconnected ones wherein selection occurs. The limiting behavior is generally deterministic. Yet, since there are selective effects on randomness in the history of lineages, probability theory is shown to be a key factor in understanding the results. Besides, randomness can be maintained in the limiting dynamics, for instance to model rare mutations fixing in the population.