An analytical-plus-simulation model of a pathogen whose virulence (encoded as its mortality mean-time) mutates continuously while it spreads over a contact network. Extending network SIR/SIS dynamics with a mutating parameter yields the evolution equation for the virulence distribution and the spreading-threshold conditions, and shows that natural selection drives the population toward less-virulent strains -- because slower-killing variants keep hosts infectious longer and reach more contacts. The framework reproduces the observed succession of COVID-19 variants when fed real epidemiological data, and clarifies how different network structures modulate the effect.
Ivry, Cohen and Patron model how a continuously-mutating pathogen evolves as it spreads over a networked population. Starting from network (beyond-mean-field) versions of the classical SIR/SIS compar…