| Title: | AI-based integrative risk scores for predictive medicine and precision therapy in complex and rare diseases |
| Lead Institution: | Queen Mary University of London |
| Principal investigator: |
Application 96802
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
About
The evolution of the artificial intelligence (AI) is promising significant advances in the area of health science and health care. With our ability to harvest a significant amount of information from large populations, like UK Biobank, we are extending our horizons on better prevention, diagnosis and treatment of diseases. AI methods would allow us to better process and combine multiple layers of health-related information, including genetic predisposition, gene expression, clinical and biochemical characteristics, lifestyle and behavioral factors, in order to customize the way we predict the disease risk at an individual level, personalize the prevention strategies and offer tailored treatment options that maximize the therapeutic effect. We hypothesize that by using sophisticated analytical methods and a range of potential predictors for a disease, we can create novel, high-accuracy prediction tools that could be easily implemented in the clinical practice. To that aim, we propose to utilize the deep phenotyping and multi-modal measurements of the UK Biobank resource to develop, test and fine-tune these AI-based integrated risk assessment tools. We aspire that the adoption of such novel integrated risk scores in clinical practice, will be beneficial for reducing NHS costs, both due to more effective disease prevention and disease management. To comprehensively assess how such tools might be accurate across different diseases, we have selected to investigate a range of high burden diseases, including the cardiovascular disease spectrum, type 2 diabetes, obesity, familial hypercholesterolemia, fatty liver disease, thyroid disease, frailty and COVID-19. We will be focusing both on common and rare sub-phenotypes of these diseases, to further fine-tune the utility of our tools for patient care. We plan to validate our novel tools to other population cohorts and assure transferability to non-European ancestry groups. Finally, we plan to test these AI-based integrated risk scores in clinical trial settings. We anticipate our project to last for 3 years at minimum.1 Publication
| Pub ID | Title | Author(s) | Year | Journal |
|---|---|---|---|---|
| 21430 | Genome-wide association analyses of borderline personality disorder identify 11 loci and highlight shared risk with mental and somatic disorders | Fabian Streit (+150) | 2026 | Nature Genetics |
Enabling scientific discoveries that improve human health