Participant identification accuracy by combining subnetworks. For the geodesic distance, the frontoparietal (subnet1) and default mode (subnet2) subnetworks were combined. For the Pearson dissimilarity measure, the dorsal attention (subnet1) and default mode (subnet2) subnetworks were combined (the top two subnetworks based on mean accuracy across conditions for this measure). Abbreviations: EM, emotion processing; GB, gambling; LG, language; MT, motor; RL, relational processing; RS, resting-state; SO, social cognition; WM, working memory.