Dependence of the control energy on the number of driver and target nodes.

(A) Energy to control the whole brain network (median over subjects) as a function of the number of driver nodes n_d. For each n_d, nodes were selected based on a ranking of centrality measures. (B) Energy to control the whole brain network (distribution over subjects), for three values of n_d. For each subject, energy values were z-scored with respect to the mean of the random node selection. (C) Energy to control a varying number of target nodes, using n_d = 5 driver nodes selected according to different centrality measures as well as randomly. Lines represent the average control energy over subjects (over both subjects and realizations for the random curve) (D) Energy to control target nodes, using n_d = 10 driver nodes (distribution over subjects). For each number of target nodes, energy values were z-scored with respect to the mean energy obtained with the random node selection. (E) Energy to control a varying number of target nodes within each of 8 RSNs, using n_d = 10 driver nodes selected according to a ranking based on the out-strength. In (A-D), all random curves were obtained by averaging over M = 100 random selections of driver nodes.