Research Paper Volume 16, Issue 9 pp 8361—8377
Comprehensive bioinformatics analytics and in vivo validation reveal SLC31A1 as an emerging diagnostic biomarker for acute myocardial infarction
Figure 4. Evaluation of the stacking-based AI predictor for AMI diagnosis and comparison of its efficacy with individual genes involved (i.e., PDHB, CDKN2A, GLS, and SLC31A1) and gold standard biomarkers (i.e., TNNI3 and CKM). (A–C) ROC curve of the stacking-based AI predictor in the training set, validation set, and test set, respectively. (D–G) ROC curve of PDHB, CDKN2A, GLS, and SLC31A1, respectively. (H, I) ROC curve of TNNI3 and CKM, respectively.