Figure 1: Overlap of the detections between the methods. The Venn diagrams illustrate the numbers and overlaps of the detections (FDR<0.05) made by the ordinary T-test, the Rank Product (RP), the SAM method, or the reproducibility-optimized test statistic (ROTS) in the different comparisons: (A) T1D vs NGT, (B) T1D vs NGT+T2D and (C) T1D vs T2D. LIMMA and Wilcoxon rank sum test detected almost nothing and they were excluded from here for the clarity of illustration. The numbers of detections with all the methods are shown in Table 2.