Research Topic: Deep Learning from Primary Samples to Predict Treatment Response in Glioblastoma: the Deadliest Adult Brain Cancer
Supervisors: Dr Lucy Stead, Prof Andy Bulpitt
About Morgan: Originally from South Wales, I studied a BSc in Biomedical Sciences at the University of Bath with a years placement in a molecular biology research lab. I then studied an MRes in Cancer Biology at Imperial College London, during which one of my projects involved applying machine learning to predict cases of endometrial cancer using omics data.
Project description: Glioblastoma (GBM) is an incurable brain cancer, with almost half of patients dying within a year of diagnosis. This dismal prognosis is due to the rapid recurrence of chemotherapy and radiotherapy-resistant GBM following surgical resection of the primary tumour. We have observed that recurrent tumours exhibit up-regulation or down-regulation of a particular set of genes. Importantly, the direction of these gene expression changes appears to influence the mechanisms by which tumour cells confer therapy resistance. Thus, my project aims to employ machine learning to predict whether the recurrent tumour will up-regulate or down-regulate this set of genes using data from the primary tumour. This could allow us to tailor GBM patients’ therapy to treat their disease more effectively.