Multi-omics for multi-morbidity
Supervisors:
Prof Evropi Theodoratou, Usher Institute (School of Population Health Sciences), University of Edinburgh
Prof Saturnino Luz, Usher Institute (School of Population Health Sciences), University of Edinburgh
Prof Jackie Price, Usher Institute (School of Population Health Sciences), University of Edinburgh
Summary:
This project investigates how multimorbidity disease clusters can be understood through multi-omic profiling and longitudinal health data to uncover shared biological mechanisms. Using the large population cohorts UK Biobank and Generation Scotland, firstly we aim to identify novel disease clusters based on disease prevalence, incidence, and temporal disease trajectories. Conventional statistical approaches, including k-means and hierarchical clustering, will be combined with advanced machine learning techniques such as graph clustering to detect patterns of disease co-occurrence and potential shared causes.
The second aim focuses on characterising the multi-omic profiles of the most promising disease clusters. Differences in demographic, clinical, and environmental risk factors will be analysed using statistical tests and regression models, while advanced methods such as dimension reduction, mixture modelling, and multi-omics network analysis will help uncover underlying biological pathways. The project also involves active Patient and Public Involvement (PPI) to improve research relevance and delivery. Students will develop expertise in biomedical data science, statistical modelling, machine learning, health data analysis in R, and the interpretation and communication of complex biomedical findings.