Unravelling resistance evolution using liquid biopsies

A major challenge in medicine, especially cancer treatment is the emergence of resistance to therapy.

Blood-based, or liquid, biopsies present an unprecedented opportunity to gain new insights into resistance evolution, but current techniques are not suitable for their analysis.In this project, we uncover the dynamics of resistance during therapy, using a combination of bioinformatic, machine learning and mathematical modelling techniques applied to next generation-sequenced liquid biopsies from cancer patients.

We (Aim 1) establish bioinformatic methods to enhance the disease-specific signal gained from liquid biopsies; (Aim 2) build a hybrid framework combining a neural network and a mechanistic model to infer cancer dynamics from time-resolved sequencing data; and (Aim 3) identify patient-specific resistance dynamics, biomarkers and optimised sampling schedules from regularly collected patient samples.

The project is carried out by the main applicant (50%) and a PhD student (100%) and is scheduled for 2025-2028: the three aims are to be completed in 2025, 2027 and 2028, respectively.This project will shed light on the timing, tempo, and molecular mechanisms of how drug resistance emerges during cancer therapy.

The tools developed here will open a new avenue in personalised medicine harnessing the full potential of cheap and non-invasive liquid biopsies to optimise monitoring and treatment.