Knowledge graph based AI for multi-omics data in precision oncology

Please apply through 'INSTITUTE WEBLINK LINK'

This Engineering Doctorate (EngD) position will work within a long-term collaboration between bioinformaticians within Canon Medical Research Europe’s AI Centre of Excellence (based in Leith, Edinburgh) and the AI & digital pathology unit in Glasgow University’s Institute of Cancer Sciences (ICS, based in Bearsden, Glasgow). The project will explore the use of graph convolutional neural networks (GCNs) as a tool to extract understanding from bioinformatics databases and multi-omic clinical data, with applications to precision medicine and clinical decision support. We will ask, for example:

How can knowledge graphs be extracted and combined from the multiple bioinformatics databases?

How to apply such knowledge graphs to improve prognostic predictions in cancer genomics?

How can evidence from different -omics modalities (e.g. transcriptomes, somatic genomes, immunohistochemistry) be combined and explained?

How can large, complex knowledge graphs be distilled into rules applicable in clinical decision support?

As an EngD student, the chosen candidate will spend the majority of their time working with the industrial sponsor, Canon MRE. There will be a strong connection with Glasgow University, which will provide academic supervision, training and assessment, including taught courses and graduate social & development programmes. The detailed balance between the two sites, and the timing of different components, are flexible.

Candidates should have a strong background in informatics, mathematics or physics. Good programming skills are essential, as is enthusiasm for application of machine learning to health and biomedical enterprises. Experience with Python, bioinformatics and machine learning would be an advantage.