Fig. 4

Results of the Log-rank test and the Cohen’s kappa statistic. a The p values by the Log-rank test. The x-axis is methods and the y-axis is datasets. A steel blue point means that the p value of the method is less than 0.05, which indicates that the method can successfully stratify patients in the dataset into two risk groups with significantly different survival patterns. A (small) grey point means the two risk groups are not significantly different between each other by the method on the dataset. b The histogram of concordance values estimated using the Cohen’s kappa statistic. There are six levels of concordance values between methods (More information can be found at Additional file 1)