Multiomics Models Show High Discrimination For Stroke
A systematic review (PROSPERO-registered) evaluated machine learning models using multiomics data for stroke risk stratification, identifying seven studies (n=40,274) published 2022–2025. Studies combined two omics layers and reported AUCs between 0.75 and 0.97, yet external validation and calibration were limited. Authors urge leakage-resistant evaluation, site-specific external validation, and benchmarking against single-omics and clinical baselines to improve reproducibility.
Key Points
- 1Identify seven studies (n=40,274) from 2022–2025 integrating at least two omics layers
- 2Show high apparent discrimination with AUCs 0.75–0.97 but limited external validation
- 3Recommend leakage-resistant evaluation, site-specific external validation, and benchmarking versus clinical baselines
Scoring Rationale
Comprehensive systematic review provides useful synthesis; limited by few heterogeneous primary studies and sparse external validation.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
