Deep Learning for Precision Oncology

All cancer types can potentially resist therapy, either through innate mechanisms or acquired in response to treatment. Patients with resistant tumours have poor prognosis and around 460 people die of cancer every day in the UK alone. Accordingly, better and more effective therapies are urgently needed, including: novel molecular targets, new drug indications, companion diagnostic biomarkers and approaches to enable selective immune targeting. Promising approaches towards these precision oncology tools seek to identify molecular ‘weak points’ in cancer and to understand the mechanisms that drive resistance to therapy. Indeed, rich genome-scale multimodal data are available for cancer cells and patients, with matched readouts of sensitivity to potential therapeutic inhibition of protein targets. These data are ripe for interrogation by cutting-edge AI methods, which will be applied to enable integrative drug discovery at scale. This project aims to overcome drug resistance in multiple cancers, towards longer remission times and increased likelihood of a complete pathological response (tumour clearance). Discovery of biomarkers would ultimately enable more effective prescribing across a range of clinical pathways and could lead to earlier diagnosis of drug-resistant cancers.

The successful candidate will be based in the Overton group (www.overton-lab.uk) and co-supervised by Dr Son Mai, benefiting from an industry placement and guidance from Mevox Ltd. The student will be furnished with skills and knowledge necessary to succeed as an independent research scientist, including training in: cancer biology, deep learning, polyomics data integration, cluster computing, network biology, structural bioinformatics and transferable skills (scientific method, disseminating results, computer programming etc.). Student development will be guided by assessment of specific needs.

Click here to access the application form