Genetic shifts in viral populations: how sudden vs gradual changes in host species composition affect the stability and dynamics of genetic variability in viral populations

Sudden vs gradual changes in host species composition —from more to less permissive− have been shown to affect viral fitness, virulence, and rates of molecular evolution. In this study, we characterized the role of noise and selection in previously published time series data from evolving populations of Sindbis virus under different rates of host replacement. First, an approximate Markov model within the Wright-Fisher diffusion analytical framework indicated a reduction in effective population size by about half, along with the co-occurrence of fewer and generally weaker beneficial mutations in the scenario of sudden host composition change. Second, analysis of genetic distance between populations at consecutive time points showed that populations under gradual host change evolved steadily until the original host was no longer present, whereas distances to the ancestral sequence showed evidence of leapfrog phenomenon at some points. In contrast, the sudden treatment was associated with relatively less stable population compositions that diverged from the ancestral sequence at a more constant rate. Third, we found that allele frequencies and their dispersion follow a power law known as Taylor’s law. Both treatments showed a high degree of allele aggregation with large fluctuations, where neutral, beneficial and deleterious alleles can be discerned by their general behavior and the region they occupy on Taylor’s plot. Finally, we found evidence that host replacement regime affects how the mutations are distributed across the genome by time.