About
My research focuses on developing individual-level, life-course models of disease risk. These models enable us to investigate how gene–environment interactions influence health outcomes and to identify critical periods across the life course during which genetic effects on disease emerge or change. To support this work, we create novel biostatistical methods, including the BGM package, and bioinformatics tools, such as ADPr-Seq. In collaboration with our research partners, we are also developing new genomic assays for the identification of DNA modifications. Our methods and software are available through my personal and research-group GitLab repositories, while associated datasets are deposited in Figshare.
Our methodological and applied research has contributed to important discoveries in complex disease genetics, including the identification of genetic factors associated with body mass index and the characterisation of shared genetic architecture across immune-mediated diseases. One particularly notable study identified distinct genetic influences on BMI at different stages of life, demonstrating how its genetic architecture changes from infancy to adulthood. This work has helped to elucidate the molecular mechanisms through which genetic variation affects disease risk across the life course.
Our research on adipose tissue has identified disease-relevant, cell-type-specific expression quantitative trait loci (eQTLs). We have also developed models to disentangle the effects of correlated environmental and physiological factors, such as circadian rhythms and fasting, on gene regulation and to characterise the corresponding context-specific eQTLs. These findings provide insight into how disease-associated genetic variants exert regulatory effects that vary across cell types, tissues and physiological conditions. Collectively, this work has important implications for precision medicine, particularly for understanding metabolic disease mechanisms and identifying and prioritising potential therapeutic targets.
Machine learning and artificial intelligence are integral to our interdisciplinary approach. I co-developed GRAD, a machine-learning method for feature selection, and introduced methodological advances in Bayesian model averaging for predicting clinically relevant outcomes, including sepsis prognosis. I have also developed Bayesian methods for analysing longitudinal data, as well as inductive logic programming approaches for modelling time-series data.
By integrating statistical theory, computational methodology and biological application, we develop tools that extract clinically and biologically meaningful insights from complex genomic and health data, including insights that may not be accessible using conventional analytical approaches.