Group x Parcel interaction effect for cerebellar and striatal connectivity. Parallel analyses were conducted independently for the cerebellum and striatum, using the functionally-defined 7 cerebellar and 6 striatal parcels. (a) Surface view (left) and volume slice view (right) of regions of interest (ROI) revealing a significant Group x Parcel interaction effect for the cerebellar analysis. Slice outlined in yellow is magnified in inset in panel b. (b) Each of the 37 cerebellar interaction ROIs were assigned to 1 of 3 sets based on the between-group differences in functional connectivity of all 7 cerebellar functional parcels using an average-linkage hierarchical clustering algorithm, such that ROIs within a set share the most similar patterns of group differences in seed-based connectivity. Inset shows location of ROIs color-coded for each set. (c) To illustrate the sources of the interaction effect, the mean functional connectivity for each set of ROIs is shown separately for SCZ and CON subjects in each of the cerebellar parcels, averaged across all ROIs in the set. Error bars indicate standard error. ROIs in Set 1 exhibit higher mean connectivity in patients compared with controls, whereas ROIs in Set 2 exhibit lower mean connectivity in patients. (d) This effect is more pronounced for associative parcels (pink) than for sensory parcels (teal). ROIs in Set 3 are less differentiated between patients and controls. (e–h) Parallel analysis conducted for the 60 interaction ROIs from 6 striatal functional parcels. (i–k) Correlations between mean functional connectivity (Fz) of all cerebellar and striatal parcels in (i) Set 1, (j) Set 2, and (k) Set 3 interaction ROIs, across SCZ and CON subjects. Displayed r values are across all subjects (N = 320, 1 outlier removed). Abbreviations: Sens., sensory networks; Assoc., associative networks; VIS, visual network; SOM, somatomotor network; DAN, dorsal attention network; VAN ventral attention network; LIM, limbic network; FPCN, frontoparietal control network; DMN, default mode network.