About the project
Why do people with similar genetic risk develop different health outcomes, and can AI identify who is most at risk? This project will develop uncertainty-aware statistical and machine learning methods for personalised disease risk prediction, combining genetic information, clinical measurements and potentially longitudinal health data.
Genetic information and clinical measurements offer new opportunities for personalised disease risk prediction. However, individuals with similar genetic profiles may experience very different health outcomes. AI models often overlook this variability and the uncertainty arising from incomplete data, measurement errors and differences between populations. This can lead to unreliable or overconfident predictions.
This project will develop statistical and machine learning AI methods to improve personalised risk prediction by explicitly representing and quantifying uncertainty in genetic and clinical data. Your research will address three interconnected challenges:
- multimodal risk prediction: Integrating genetic, clinical and imaging-derived biomarkers to develop personalised risk models
- uncertainty quantification: Developing hierarchical models that distinguish different sources of uncertainty and assess the reliability of individual predictions
- reliable decision support: Using computer simulations and real-world data to evaluate prediction accuracy, bias, calibration and uncertainty, with particular attention to clinical decision-making
The primary application will focus on metabolic diseases, particularly type 2 diabetes and obesity-related complications. The project will investigate how genetic predisposition, clinical biomarkers and their changes over time contribute to disease risk, using suitable population-based and clinical datasets. You'll work at the intersection of statistical methodology, genetic epidemiology and trustworthy AI, developing methods that move beyond accurate population-level predictions towards reliable and clinically meaningful individual risk assessments.
You'll receive interdisciplinary training in statistical modelling, machine learning, genetic epidemiology and uncertainty quantification. Training will include analysing large-scale genetic and longitudinal clinical datasets, designing simulation studies, assessing prediction reliability and bias, and applying reproducible research practices. Supervisors will support scientific writing, conference presentations, collaboration and career development.
The School of Electronics and Computer Science is committed to promoting equality, diversity inclusivity as demonstrated by our Athena SWAN award. We welcome all applicants regardless of their gender, ethnicity, disability, sexual orientation or age, and will give full consideration to applicants seeking flexible working patterns and those who have taken a career break. The University has a generous maternity policy, onsite childcare facilities, and offers a range of benefits to help ensure employees’ well-being and work-life balance. The University of Southampton is committed to sustainability and has been awarded the Platinum EcoAward.
