Predicting and steering the evolution of multidrug resistance in Pseudomonas aeruginosa

Evolution of antibiotic resistance is inevitable during long-term treatment and contributes to deaths during chronic infections. However, bacteria can only adapt to current conditions.

If we could steer their evolution, it might be possible to trick them into evolutionary dead ends where they can no longer evolve resistance or do so only at a cost of reduced growth or ability to cause disease.

This project will develop a framework for predicting and steering the evolution of multidrug resistance in the opportunistic pathogen Pseudomonas aeruginosa using a combination of mathematical modeling, experimental evolution and genomics.

In this three-year project, an interdisciplinary team of mathematical and experimental biologists will work to combine experimental data with mechanistic mathematical models of the molecular networks that drive resistance evolution to predict the evolution of bacterial populations in the laboratory.

Later, focus will be on guiding bacteria to desirable genetic outcomes by predicting the environmental conditions likely to lead to a certain evolutionary trajectory. Finally, the large amount of phenotypic and genomic data will be used to develop machine learning models.

Use of clinically relevant antibiotics means that results can inform antibiotic treatment protocols to minimize resistance and increase treatment success in patients and the framework can be applied to other bacteria that evolve resistance primarily by chromosomal mutations.