Figure 2 | Scientific Reports

Figure 2

From: Identification and validation of BCL6 and VEGFA as biomarkers and ageing patterns correlating with immune infiltrates in OA progression

Figure 2

Construction of OA risk predictive model using the SVM and RF methods. Boxplot of the residual distribution (A) and reverse cumulative distribution of residual (B) as a function of the values and ROC curves (C) showing the observed sensitivity between RF and SVM. (D) RF: prediction error curves based on tenfold cross-validation. (E) The importance of the 30 senescence regulators based on the RF model. (F) Nomogram graph of the predictive model based on two m6A regulators. The calibration curves (G), clinical impact plot (H) and DCA (I) were used to determine the clinical utility of risk prediction nomograms. SVM, support vector machine; RF, random forest; OA, osteoarthritis; ROC, receiver operating characteristic; DCA, decision curve analysis.

Back to article page