Supplementary Figure 2
(A) For the datasets as indicated by their respective labeling, it is assessed whether the complex median correlation (Pearson’s r) is biased by the abundance of the respective complex (first row, grey shading) or by the complex variance (second row, blue shading). For comparability, abundances and variances are rank-sorted and further split into 25%-bins; the median correlation is then monitored in each bin. While they were significant differences between some bins (t-test, < 0.1 (*), < 0.05 (**), < 0.01(***)), no general trend could be observed and also those significances could not be recovered consistently across datasets.
(B) For the same datasets as above, median correlation values (Pearson’s r) were monitored for randomly assembled complexes (decoy complexes) from permuted data (reshuffled data) (grey, first boxplot), decoy complexes from original data (light-blue), and real complex sets from original data (purple).
This figure has been generated with the following code: suppFigure2_code
The underlying data can be found here.