Figure 2
From: New network topology approaches reveal differential correlation patterns in breast cancer
Dependence of the resulting gene list on the cutoff point. (A) Algorithm DCglob: Number of detected genes in dependence of the cutoff point for the change of global network topology. The parameter p is connected with the length of the interval where the change of correlation is stable. Additionally, the mean number of detected genes comparing 100 pairs of breast cancer random subsamples including a 90% confidence interval is shown. (B) Algorithm DCloc: Number of detected genes in dependence of the cutoff point for the changes of local network topology. The parameter d can be interpreted as proportion of correlated genes that are not shared by condition A and B. Similarly, the number of detected genes comparing 100 pairs of breast cancer random subsamples including a 90% confidence interval is shown. (C) Comparison of DCglob and DCloc: The number of detected genes in dependence of the FDR.