FIGURE 4.
Separation of benign and malignant patients by machine learning of proteomic features. (A) Top 5 proteins prioritized by random forest analysis ranked by the mean decrease in accuracy > 5. (B) Receiver operating characteristic (ROC) analysis of the classifier and each feature in the training dataset. (C) Expression levels of the five proteins; p-value was calculated in t-test medtod. *B, benign group; M, malignant group.