Skip to main content
. 2020 Sep 2;40(36):6949–6968. doi: 10.1523/JNEUROSCI.2559-19.2020

Figure 12.

Figure 12.

Reassignments conferred by the variants of network partition deviation in the control cartographies. Reassignment is the network index (based on the intrinsic partition) of the highest mean connectivity estimate, per task state. In each panel, Jaccard indices are listed to indicate similarity between two partitions. Network assignments are color coded (as in Fig. 3), and the 64 C-PRO task states are collapsed into 12 rule sets (plus Rest as a reference on each x-axis). A, left, Network reassignments of CON regions from the deviation algorithm, with all connectivity data included in the input (Fig. 11A, x-axis). Middle, Within-network CON estimates used in the deviation algorithm (Fig. 11C, y-axis; see Fig. 5C). The Jaccard similarity of within-network and all-data deviation is 0.31. Right, Between-network CON estimates used in the deviation algorithm (Fig. 11C, x-axis; see Fig. 5D). The Jaccard similarity of between-network and all-data deviation is 0.16, which is lower than the within-network similarity to all-data. B, Same as panel A, except for FPN regions. The Jaccard similarity scores are: within-to-all = 0.18, between-to-all = 0.29. Thus, FPN showed the reverse pattern to CON, where between-network deviation is more similar to all-data deviation than within-network deviation.