Figure 1. . Motifs, motif fingerprints, and motif morphospace. (A) All six possible four-node undirected motifs. Motif nodes are colored according to their degree, where yellow nodes have degree 1, orange have degree 2, and red have degree 3. Motif edges forming triangles are highlighted in bold. (B) Motif fingerprint averaged across all subjects (blue). The error bars show ±σ, where σ is the standard deviation. The motif fingerprint from three different null models are shown for comparison, namely randomized networks not preserving the degree distribution (purple), randomized networks preserving the degree distribution (red), and randomized networks obtained by scrambling the wavelet coefficients (orange). In the last two cases results were averaged over 100 realizations per real fMRI network. (C) The motif PC morphospace plotted using the first two PCs. Here each red dot represents an individual subject, plotted in PC space. The original motifs are also plotted as vectors in the PC space. The axes are labeled as small-worldness and disassortativity due to the correlations of PC1 and PC2 with global network metrics.