TRACE: An Integrative Liquid Biopsy Framework for Monitoring PARP Inhibitor Resistance

PARP inhibitors (PARPi) are an important treatment for cancers with defects in DNA repair, including many ovarian, breast, prostate, and pancreatic cancers. However, cancers can evolve during treatment and eventually become resistant. Detecting this resistance early is challenging: current monitoring relies largely on imaging, while repeated tumour biopsies are invasive and often impractical. Blood samples offer a minimally invasive way to track how a cancer changes during treatment.

This PhD will develop TRACE, a computational liquid-biopsy based framework for detecting emerging PARPi resistance using whole-genome sequencing of tumour-derived DNA circulating in the blood. Rather than relying on a single resistance marker, TRACE will combine multiple types of genomic information and use established resistance mechanisms observed in tumours to guide their interpretation in blood. The overarching aim is to develop and independently evaluate a blood-based approach for more sensitive and timely monitoring of treatment resistance.

Year 1: Using an existing ovarian cancer cohort with tumour WGS and longitudinal blood samples per patient, the student will identify genomic features associated with PARPi resistance and determine how they appear and evolve in circulating DNA. This will include alterations in DNA-repair genes, structural and copy-number changes, mutational patterns, and other genome-wide features.
Year 2: The student will develop TRACE, applying machine-learning approaches to integrate complementary genomic signals and identify those most informative for detecting emerging resistance.
Year 3: TRACE will be independently evaluated using an independent patient cohort, testing its accuracy and whether resistance can be detected earlier than with existing biomarkers or clinical monitoring.
Year 4: The framework will be developed into a reproducible, open-source research tool and prepared for evaluation in other PARPi-treated cancers.

The student will gain interdisciplinary training in cancer biology, liquid biopsy, WGS analysis, bioinformatics, statistics and machine learning, explainable AI, and reproducible computational research.