| Title: | Development and validation of a comprehensive predictive model for diabetes complications |
| Lead Institution: | Catholic University of Korea |
| Principal investigator: | Professor Jae-Seung Yun |
Application 67855
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
Aims: As diabetes prevalence is increasing, so too is the risk of diabetic complications. This makes it very important in diabetes care to predict those at high risk for complications. To this end, we will endeavor to construct a predictive model that identifies such patients, the better to ensure they receive proper therapeutic care. Project contents and scientific rationale: Any one person's risk of diabetic complications is a composite of many contributing genetic, clinical, and environmental factors. Taking each of these aspects into account, we will perform the following analyses using data from the UK Biobank: First, we will develop a score describing the genetic component of risk for diabetic complications. Because the genetic variants that are both known to contribute to diabetes and common in the population have only modest effects on any individual patient's disease, we will use a polygenic risk score that incorporates multiple contributing genes in a single metric. We expect this score to be very helpful in classifying patients according to their degree of risk. Second, we will develop a means for predicting the progression of diabetes complications from images of a patient's retina. The retina is notable for being the only part of the body in which capillaries can be directly and non-invasively photographed. Analyzing images of retinal capillaries in patients with diabetes should enable the detection of early changes in or damage to blood vessels, thus allowing early prediction of and prompt response to diabetic complications as (and even before) they develop. Third, we plan to construct a predictive model that combines all available data from the UK Biobank. This final model will build upon the first two and take into account all three dimensions of the risk for diabetes complications - genetic, clinical, and environmental factors. Project duration: We expect this project will take about three years to complete. Public health impact: We expect that leveraging the genomic, clinical, behavioral, and socio-economic data in the UK Biobank will allow us to establish an accurate predictive model that can precisely classify those at high risk for diabetic complications, allowing for earlier preventive treatment and the most appropriate care. In addition, we expect our comprehensive predictive model will provide new insights into the natural course of diabetes progression.4 Publications
| Pub ID | Title | Author(s) | Year | Journal |
|---|---|---|---|---|
| 7210 | A deep learning model for screening type 2 diabetes from retinal photographs | Jae-Seung Yun (+8) | 2022 | Nutrition Metabolism and Cardiovascular Diseases |
| 10293 | Associations of polyneuropathy with risk of all-cause and cardiovascular mortality, cardiovascular disease events stratified by diabetes status | Kyuho Kim (+4) | 2023 | Journal of Diabetes Investigation |
| 10310 | Impacts of gender and lifestyle on the association between depressive symptoms and cardiovascular disease risk in the UK Biobank | Su Nam Lee (+9) | 2023 | Scientific Reports |
| 7216 | netCRS: Network-based comorbidity risk score for prediction of myocardial infarction using biobank-scaled PheWAS data. | Yonghyun Nam (+6) | 2022 | Biocomputing |
Enabling scientific discoveries that improve human health