I am a research fellow at the Usher Institute of Population Health Sciences and Informatics, at the University of Edinburgh. I develop machine learning and statistical methods with applications in genetic and molecular epidemiology.
My main research area is precision medicine. The goal is to identify sub-groups within the general population or a heterogeneous disease population (e.g. based on genetics or biomarker signatures) that are distinguishable with respect to disease pathogenesis, risk of disease progression, or response to treatment.
I currently work on three disease outcomes:
- Prediction of treatment response in rheumatoid arthritis : early effective therapy is consistently shown to improve long-term outcomes in rheumatoid arthritis, but all classes of drugs have substantial non-response rates (30-40%). I work with Paul McKeigue and Marco Colombo on two collaborative projects: the MRC-funded MATURA consortium and the Scottish Early Rheumatoid Arthritis cohort.
- Risk stratification for colorectal cancer: risk stratification and development of novel diagnostics are needed to increase the sensitivity of the screening programme and care pathways within it (e.g. colonoscopy follow-up). I work with Evropi Theodoratou on genetic, biomarker and environmental risk factors for colorectal cancer, analysing data from UK Biobank and colorectal cancer studies.
- Risk stratification for type 1 diabetes and its complications:it is now recognised that type 1 diabetes --previously seen as a singe disease-- is heterogeneous, with some people retaining residual capacity to secrete insulin. I work with Helen Colhoun on the genetics of type 1 diabetes and diabetic complications in the Scottish Diabetes Research Network Type 1 Bioresource Study.