Predicting the future spread of antibiotic resistance genes, has been granted funding

Research Project, 2024 – 2029

Bacteria become resistant to antibiotics through changes in their genome, often from the acquisition of antibiotic resistance genes (ARGs). The constant transfer of new ARGs into pathogens poses a significant threat to the efficacy of current and future antibiotics. Despite this, we lack fundamental knowledge about the many ARGs maintained by bacterial communities, the factors that influence their transfer into pathogens, and in what environments such transfers are most likely.

The primary goal of this Ph.D. project is to fill these knowledge gaps and clarify how ARGs are recruited and transferred into pathogens. The project will make extensive use of genomic, metagenomic, and clinical data to describe the dynamics of antibiotic-resistant bacteria. Deep learning-based methods, similar to those used in large language models, will be developed and used to identify the many previously uncharacterized ARGs by screening large volumes of sequence data. Machine learning will be utilized to analyze the factors influencing ARG horizontal transfer and to predict their potential bacterial hosts. Large-scale analysis of metagenomic data will be used to map the spread of ARGs and identify high-risk environments for their recruitment. These novel methodologies will, finally, be combined into an early warning system that identifies new emerging ARGs, evaluate the risk of their spread to pathogens, and predict their potential clinical impact.

This Ph.D. project will not only shed light on the complex evolutionary mechanisms promoting multiresistant pathogens but also establish versatile AI/ML methodologies applicable across microbiology. The project will also generate valuable data on ARGs and their spread, which will be beneficial to the infectious disease research community.