Skip to main content
Figure 4 | BMC Cancer

Figure 4

From: Robust assignment of cancer subtypes from expression data using a uni-variate gene expression average as classifier

Figure 4

Area under the Curve (AUC) values of the ranked single genes and of the metagene obtained from taking the mean expression of the single genes. A) AUC for association to MI. B) AUC for association to grade. Classification performance of MI status for the metagene based predictor (ROC) and the SVM using LOOCV in the SanchezC (C) and Stransky data (D), respectively. For SVM we used two different features selection criteria, t-test and AUC. Accuracy = Balanced Accuracy (see Methods), dashed line = Balanced Accuracy obtained by random class assignment (= 0.5). ROC = metagene-based predictor. SVM = Support Vector Machine.

Back to article page