Improving cancer therapy through evolutionary modelling

Cancer develops through a complex evolutionary process where cells acquire mutations that accelerate their growth, compete and co-operate with each other as a tumour grows. The evolutionary pressure ramps up when patients get treated, selecting for cancer cells that are highly adaptable and resilient. The unfortunate outcome is often therapy resistance that leads to patient relapse. It is difficult to observe the process of tumour evolution in patients. We use mathematical models to fill this gap, performing computation simulations of thousands of scenarios to understand how tumours are able to evolve resistance to cancer drugs. By varying the timing and intensity of drug doses in our models, we aim to discover dosing schemes and drug combinations that may confer longer tumour control and patient survival than current standards of care. We are also testing machine learning approaches to enable us to efficiently explore a wider range of treatment scenarios and rapidly home in on those that are most effective. This project presents a unique opportunity to work at the interface of math/stats, ML, cancer biology and medicine.