Fig. 6
From: The pyroptosis mediated biomarker pattern: an emerging diagnostic approach for Parkinson’s disease
Combined biomarkers constructed with the machine learning algorithm. Ranking the contribution of 10 biomarkers to the construction of predictive models (A). ROC analysis prediction model with random forest in test subset (B). The importance of 10 variables to prediction model (C). ROC analysis prediction model with random forest in H&Y stage ≤ 2 subsets (D) and disease course ≤ 5 years subset (E). The shaded region of the ROC curve represents the 95% confidence interval (miR, microRNA; circ, circular RNA; lnc, long non-coding RNA; ROC, receiver operating characteristic; AUC, area under the curve)