INTERROGATING MICROBIAL POPULATIONS: USING DEEP SEQUENCING TO ESTIMATE THE FITNESS COSTS OF MUTATIONS

Microbial populations are characterized by exceptionally high rates of evolution, manifested by high levels of heterogeneity in the microbial genomes. This heterogeneity enables the population to quickly adapt to new challenges such as newly administered drugs and the adaptive immune system of the host. The probability of emergence and retainment of a mutant in a population is dependent, among other factors, on the relative replicative fitness of the mutation. Here, we present a novel approach for estimating the fitness cost of every single mutation that occurs in a population of HIV genomes. Notably, this method is applicable to any microbial population.

Our approach relies on the well-established population genetic principle, which states that in large populations, the prevalence of a mutation is proportional to its fitness. Here, we use mutation frequencies obtained from independent HIV populations replicating in different patients, to estimate the fitness of every possible point mutation. We validate that our estimated fitness values agree with our expectations for different types of mutations (such as synonymous vs. non-synonymous vs. nonsense mutations). We further obtain intriguing new insights such as the negative fitness effects of CpG dinucleotides on the HIV genome, as well as the negative fitness effects of G→A mutations.

Finally, we exhibit our efforts towards obtaining highly accurate deep sequencing protocols that will allow us to measure the fitness effect of highly deleterious mutations and quantifying pre-existing drug resistance mutations in populations of replicating microbes.