Building Accurate Evolutionary Estimates for Large Genomic Data Sets and Applying These Inferences to Pathogen Surveillance

Tracing the evolutionary history of pathogen outbreaks allows researchers to develop appropriate public health interventions. For example, phylogenetic inferences have been key data informing the response to the on-going Covid-19 pandemic. I worked with researchers at the CDC to develop and test tools to rapidly infer phylogenies for large genomic data sets. I applied these new tools to understand the evolution of gonorrhea (Neisseria gonorrhoeae), a pathogen of major public health importance, which is increasing both in prevalence, and in rate of anti-microbial resistance. I found that our tools reduced program runtime and data set fragmentation while producing reliable phylogenetic estimates. I also investigated the underlying approach used by our methods to assemble genomic sequences. I found that reference choice is an important consideration when assembling sequences, as greater evolutionary distance to reference genome leads to an increase in errors. However, I found that while errors increase with evolutionary distance to reference genome, overall phylogenetic topology is largely unaffected. Finally, having shown that my original tools are reliable, I extended the methods and applied them to analyzing the evolutionary relationships of over 1,000 N. gonorrhoeae isolates in order to map gain and loss of anti-microbial resistant alleles data. Together these results demonstrate that the tools I have developed can be used to rapidly and accurately analyze genome scale data for thousands of lineages, and link those evolutionary inferences with important metadata to better inform public health interventions