Supplementary Figure 1
For all datasets considered for the ROC-analysis, the distribution of Pearson correlation coefficients for protein-protein pairs with STRING interaction score > 700 (combined score, orange), known interaction (combined score >0, yellow) and random protein-protein pairs (grey) is shown. The Mann-Whitney U-test was used to assess significance of the respective shifts (indicated in colored p-values). The number next to the dataset name indicates the reference for the respective dataset, with (1) referring to the The Cancer Genome Atlas (TCGA) publications, (2) Battle et al., 2015, (3) Chick et al., 2016, (4) Geiger et al., 2012, (5) Khan et al., 2013, (6) Wu et al., 2013, (7) Williams et al., 2016, and (8) Guo et al., 2015 (same numbering as for Figure 2). More details on the datasets are given in Table S1.
This figure has been generated with the following code: suppFigure1_code
The underlying data can be found here.