26 April 2019
William Hanage
Department of Epidemiology
Harvard School of Public Health
The population genetics of bacteria has received less attention than that of sexually reproducing eukaryotes, despite the fact that they do not evolve wholly clonally. Bacterial genomes are like Mr Potato Heads: they vary in gene content, sometimes to a large degree. This means that all members of a species, in the same population, will not have the same homologous loci and the genome is divided into a 'core' (the potato) and the 'accessory' (the other bits). Models that seek to explain the genomic variation we see in natural populations have paid less attention to the accessory than the core, and substantial amounts of what we see can be explained with neutral theories. However studies of different populations of Pneumococcus have shown that while their core genomes can differ to the extent that the frequency of a lineage in one place is essentially uncorrelated with its frequency anywhere else, the opposite is true of the accessory genes. The proportions of genomes from the population in any location that have any particular accessory gene are highly correlated — a finding that might be explained by Negative Frequency Dependent Selection. We will discuss how this can be used to estimate the fitness of any given strain, in any population, and describe how this has been successfully tested using data from vaccine campaigns. We will also discuss what this might mean, and how it can further inform the study of the evolution of bacterial pathogens.