Fig. 4

Receiver operating characteristic (ROC) curves illustrating the ability of miRNAS within EVs to distinguish pregnant women at early gestation who will develop GDM later during pregnancy. The predicted likelihood of woman with GDM compared to normal glucose tolerance test at early gestation. A–I Receiver-Operating Characteristic (ROC) Curves, and the area under the ROC curves (AUC) were calculated using GraphPad Prism (GraphPad Software, Inc., CA, US). J–K Classification model based on the quantification of EVs miRNAs biomarkers and developed using LogitBoost regression analysis. AUC = 1.0. At a posterior predictive probability threshold of 0.5; sensitivity = 100%, specificity = 100%. J The predicted likelihood (posterior predictive probability value) that a woman in first trimester of pregnancy will subsequently develop GDM. K ROC and AUC from J. L Performance of classification models based on the expression of individual miRNAs within sEVs, developed using a machine learning LogitBoost classifier in an external validation dataset. PPV = positive predictive values; NPV = negative predictive values