MRC DTP in Precision Medicine

Investigating the Prognostic and Predictive Values of Spatial Cellular Interactions in Colorectal Cancer through Spatial Data Analysis and Machine Learning

Supervisors

Xiao Fu, School of Cancer Services, University of Glasgow 

Joanne Edwards, School of Cancer Services, University of Glasgow 

Nigel Jamieson, School of Cancer Services, University of Glasgow 

Peter Bankhead, Institute of Genetics and Cancer, University of Edinburgh

 

Summary 

Colorectal cancer is a leading cause of cancer-related mortality. The challenge of its clinical management is influenced by the remarkable complexity of the tumour microenvironment (TME), which underpins disease progression and therapy resistance. Our previous analyses have demonstrated the prognostic and predictive value of peritumoural inflammation and tumour stroma percentage in a large retrospective cohort of colorectal cancer patients. Recently, we have generated multiplex immunofluorescence (mIF) data from samples within the same cohort.

This project seeks to perform more granular quantitative analyses of spatial cellular interactions using mIF data, for identifying novel biomarkers of colorectal cancer and ultimately aiding its clinical management. One arm of the project will engineer diverse quantitative features to describe spatial cellular interactions and evaluate their association with clinical outcomes. The other arm of the project will develop graph-based deep learning models to predict survival and cancer recurrence.

This interdisciplinary project is an exciting opportunity for a student passionate about developing new skills and applying quantitative data analysis and machine learning to spatial cancer datasets with comprehensive clinical and pathological annotations. The successful candidate will work alongside a diverse team of computational scientists, bioinformaticians, clinicians, and experimentalists in the highly dynamic and collaborative research environment.