Predicting Ovarian/Breast Cancer Pathogenic Risks of Human BRCA1 Gene Variants of Unknown Significance

Figure 1

ROCs of 4 kinds of predictive models. ROCs indicating varied performances of different models were plotted. And the relevant AUCs were also computed to indicate models’ overall performance. For support vector machine (the light blue curve), random forest (the purple curve), PolyPhen (the red curve), and SIFT (the green curve), their AUC values were 0.74, 0.78, 0. 74, and 0.78, respectively.