INFERRING THE FITNESS OF MUTATIONS IN MICROBES

Microbes evolve at exceptionally high rates, allowing them to adapt to a multitude of environments. In order to understand and predict the course of evolution of microbes, it is critical to know the how mutations affect the fitness of the microbial population. Today, with the advent of accurate next generation sequencing techniques, there are now available time-series data of mutation frequencies present within a microbial population. Here we present a method based on Approximate Bayesian Computation (ABC) that enables inferring the fitness values of mutations based on the changes in mutation frequency across time in the population.

We used this method to study the reversion of the attenuated oral poliovirus vaccine strain 2 back to the virulent, wild-type phenotype, and identified mutations under positive selection when the virus is grown at high temperature. Our analysis corroborated known attenuating mutations, in addition to identifying putative novel sites. Identifying these mutations as advantageous or deleterious to the evolving virus may help to design better vaccine strains and aid in forecasting microbial evolution.