Fig. 4.

Evaluation of network analysis methods within co-expression modules by WGCNA on GNW3000 networks a Barplots of average pAUPR for different methods. Error bars represent standard errors of the pAUPR values across the top 1000 predictions. A Student t-test was carried out to determine the significance of the difference of pAUPR values between RLowPC and the other eight methods. P-values are shown on the top of the bars if it is less than 0. b Box plots of precisions in different groups of top 1000 edge predictions. The means of precision within modules by WGCNA (0.0057) and before clustering using WGCNA (0.0031) are shown as red and blue dashed lines