Fig. 2From: A comparative study of survival models for breast cancer prognostication revisited: the benefits of multi-gene modelsThe distribution of biomarker performance on each of the three models tested in four selected meta-dataset partitions. GeneSIMMS biomarkers are separated according to size and the performance of the single-gene AURKA model is displayed. Highlighted datasets comprise the testing cohort in each partition. The total number of patients in the testing cohort is also provided. The plots for the remaining eight partitions can be found in Additional file 1: Figure S1Back to article page