Proteomics‐enabled learning machine algorithms enhance the prediction of cardiovascular diseases in patients with type 2 diabetes mellitus
Authors' conclusion
Proteomics can be used to predict cardiovascular complications in diabetic patients. It is also necessary to consider the applicability of the model due to the limitations of the sample size and the constraints of proteomics in clinical applications.
Study context
- Population
- Patients with T2D, UK Biobank, 2,920 Olink proteins
- Participants
- 874 patients
- Comparison
- Full model vs basic model
- Outcome
- Prediction of CVD, evaluated by AUC
Methodological detailsmachine learning, random survival forest, proteomics, Cox and LASSO regression for protein screening
- Machine learning
- Random survival forest
- Proteomics
- Cox and LASSO regression for protein screening
Performance
As reported in the abstract
Time-dependent AUC for predicting CVD · 3, 5 and 10 years
| Test or model | AUC 3 years | AUC 5 years | AUC 10 years |
|---|---|---|---|
| Full model | 0.81 | 0.74 | 0.74 |
| Basic model | 0.77 | 0.69 | 0.67 |
AUC 3 years
Full model0.81
Basic model0.77
AUC 5 years
Full model0.74
Basic model0.69
AUC 10 years
Full model0.74
Basic model0.67
Further findings
Comparison with Framingham, ASCVD and Score2-Diabetes models
- Prediction performance of the full model was better than Framingham, ASCVD and Score2-Diabetes models
Identification of high-risk individuals (net reclassification improvement and integrated discrimination improvement)
- Full model identified high-risk individuals more accurately; accuracy rate
- 79% vs. 69%
Original publication
Proteomics‐enabled learning machine algorithms enhance the prediction of cardiovascular diseases in patients with type 2 diabetes mellitus. Diabetes Obes Metab. Published 1 September 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →