This page will soon be decommissioned. Last update: May 2022.
Please visit my new homepage.
I am a Chancellor's Fellow in Data-Driven Innovation at the Usher Institute of Population Health Sciences and Informatics, at the University of Edinburgh. I have formal training in machine learning and ten years' research experience in statistical genetics and molecular epidemiology.
I am a member of the Molecular Epidemiology group (Usher Institute) and the Centre for Statistics (School of Mathematics) and I hold a research affiliation with the Centre for Genomic and Experimental Medicine at the Institute of Genetics and Cancer.
My research aims to address questions about the pathogenesis, progression and treatment of disease by developing and applying statistical methods to perform predictive and causal inferences. A key component of my work is the analysis of large-scale, multi-source and high-dimensional datasets, combining genetic and biomarker data (omics) with electronic or conventionally collected health records.
I have mainly focused on applications of precision medicine on autoimmune diseases. The goal is to stratify individuals based on genetic or biomarker signatures that distinguish different subtypes of disease, and to develop corresponding predictive and prognostic biomarkers. Predictive biomarkers of treatment response enable targeting of treatments to the group of patients most likely to respond. Similarly, prognostic biomarkers can enhance disease management, such as screening for complications and recruiting for clinical trials, by identifying patients with a worse disease outlook.
- The GENOSCORES platform: an online software tool for computing genotypic risk scores from summary results of genome-wide association studies and using these for prediction and causal inference.
- Prediction of treatment response in rheumatoid arthritis:
biologic therapies have revolutionised the outlook for severe
rheumatoid arthritis, but there is substantial variability in response to
treatment among rheumatoid arthritis patients for all classes of drugs.
This has spurred efforts to discover predictors of response, as early
effective therapy is consistently shown to improve long-term outcomes.
Collaborations: MATURA consortium, TRACT-RA project.
- 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. Identifying genetic
determinants and biomarkers of type 1 diabetes and diabetic complications
can help us understand the underlying pathogenic processes and better inform
decisions for preventing, managing and treating complications.
Collaborations: Helen Colhoun's group, the Scottish Diabetes Research Network Type 1 Bioresource Study.
- 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).
Collaborations: Evropi Theodoratou's group.
- Genetic factors of giant cell arteritis, its complications and
response to treatment: people with giant cell arteritis are at risk
for ischaemic complications, including irreversible vision loss.
Glucocorticoid monotherapy is the typical treatment, however, up to 50% of
patients remain glucocorticoid dependent 2-3 years later, leading to significant
toxicity and adverse events.
Accumulating evidence points to a genetic predisposition for giant cell arteritis,
but lacks robust characterization of the genetic variants contributing to disease
risk, its complications and response to treatment.
Collaborations: Ann Morgan's group, TARGET Consortium.