Evolution and emergence of novel human infections

Arinaminpathy, N.; McLean, A. R.

We now incorporate adaptation, using a modified version of a model presented by Antia et al. (2003) , based on multi-type branching processes. The biological motivation for the model is as follows: an index individual infected with a wild-type (poorly adapted) pathogen is likely to die or recover without infecting anyone else. However, owing to the possibility of the pathogen mutating to acquire some transmission potential, this index case has a non-zero probability of infecting another person. Subsequent adaptations can serve to increase the probability of transmission still further. The stochastic process continues either until extinction of the pathogen, after a certain number of cases, or until emergence of an adapted pathogen. The latter may be subject to yet further adaptations, but we do not consider these stages here. In case of extinction, the process is repeated with another new introduction of a wild-type pathogen from the animal reservoir. In this way, we simulate a series of introductions leading to emergence. Whereas Antia et al. (2003) presented a discrete-time model, ours is a continuous-time process as we are also interested in the time course of an outbreak. Moreover, as discussed below, in this model the role of R 0 is to reflect relative probabilities of transmission and recovery; in the context of continuous time, this allows a more natural description of the process of adaptation.