Our recent work on influenza reassortment has now gone live:
Experiments measuring rates of influenza reassortment on the time-scale of individual inections have typically been conducted in small animal models or in vitro. After initiating infection with distinct viral strains a measure of the rate of reassortment is obtained by examining the extent to which new combinations of genome segments have been formed after a period of time.
Here we took an alternative approach, using changes in the genetic diversity present in an influenza virus population in human infections to measure the rate of reassortment within a human host. During infection, natural selection for specific genetic variants in the virus led to changes in the frequencies of variant alleles over time. The manner by which variant frequencies change over time contains information about whether segments evolve independently of one another (as under rapid reassortment, potentially involving epistasis between variants) or in a linked manner (that is, under limited reassortment).
Via an evolutionary model we found evidence that the effective rate of reassortment (that is, between distinct viral segments) is significantly limited in a human host, being substantially less frequent than the rates previously measured. Previous studies have shown that the influenza populations within a host may evolve in genetically distinct manners (Lakdawala et al., Nature 2015). We therefore hypothesise that spatial separation between viruses leads to reassortment disproportionately occurring between viruses that are physically close in the host, so that the effective rate of reassortment is linked to local, rather than global, viral diversity. The larger size of the human airway compared to that in small mammals could increase separation between the viruses that found an infection, reducing local genetic diversity and hence also the extent of effective reassortment. The work lays ground for further research into within-host viral evolutionary processes.