Grant: Discovery Grants Program - Individual

Microorganisms are great adaptors: bacteria and other single-celled organisms see opportunity in every disrupted habitat and environmental change. Their adaptability can benefit us but it also creates new threats to us and the environment we aim to preserve. The most high-profile risk posed by microbial evolution lies in the emergence of antimicrobial resistance (AMR), a global threat that is responsible for millions of deaths per year. Our overuse of antibiotics has created an environment where resistant organisms emerge and flourish. A vital component of the microbial evolutionary toolkit is the ability to exchange genetic material between closely and distantly related organisms through processes of genetic recombination and lateral gene transfer. Despite significant progress and ever-increasing data, we still struggle to find the right ways to represent genomes and the functions they encode. Complete descriptions of transmission patterns are lacking, especially among large sets of closely related genomes where the spread of genes is most frequent. My proposed research aims to better define what is being transferred and between whom. Central to this goal is finding appropriate representations of bacterial genomes, not as bags of independent genes or biochemical pathways, but as evolutionarily coherent modules of genes that can move in concert to enable rapid microbial evolution. Mobile genetic elements are the most prominent examples, but even they change with time, and myriad patterns are less obvious. Delineation of these modules will provide much-needed context to infer major pathways of transmission and link them with key parameters such as habitat and geography. To this end, we will enhance our existing methods such as EvolCCM that can identify and cluster co-evolving genes, and use the clusters thus reconstructed as the basis for comprehensive inference of transmission based on multiple lines of evidence. Specific to AMR, we aim to better differentiate AMR genes from close variants that do not confer resistance by augmenting existing prediction methods with evidence derived from embedded representations of gene order. Doing so will increase the precision of our descriptions of the movement of AMR among members of pathogenic species. The results of this research applied to pathogenic and non-pathogenic sets of microorganisms will be disseminated via peer-reviewed journals, conferences, and open-source software. We also have specific pathways of dissemination to increase the reach and impact of our work. First, we will integrate our methods into our ARETE software pipeline that is already being used for identification of AMR transmission pathways. Second, we will integrate our gene-order-based methods into the Comprehensive Antibiotic Resistance Database, which will benefit its wide range of users. Finally, we will continue to work with our colleagues in public health to integrate our approaches with current risk models of transmission.