Fig. 6: Shared functional anatomy of data-derived networks.

a Greyscale lower-left: the fraction of shared edges between data-derived networks computed as the fraction of the smaller network contained within the larger network (e.g., 85% of the 643 edges within the AD-like (−ve) network are within the 2313 edges within the WT-aging (+ve) network). Color upper-right: Using Fisher’s exact test (corrected, permutation test ×1000), we determine if a significant number of shared edges (hot-colors), or a significant (p < 0.008) lack of shared edges (cool-colors), are found between data-derived network pairs. Number of shared edges is shown as white in each cell. b Overlap (red arrows) and lack of overlap (blue lines with caps) between data-derived networks. Arrow size proportional to significance. c Edges shared between AD-aging(+ve), WT-aging(+ve), and AD-like(−ve) pairs–green: 1–3 shown as vertical ROI plots (upper half) and histograms (lower half) with null distribution (colored histogram) and real data (gray line and black triangle) . d Bilateral regions that are identified consistently (in 1–3) with the highest degree. e Overlap between two groupings of edges shared across data-derived networks: Age, AD (left of dotted line, green: 1–3), and treatment (right of dotted line, purple: 4–6). No significant representation of edges shared across data-derived networks including the treatment network with the triple-network system. f A layered histogram of degree for the 1–3 grouping with a bird’s eye view of a degree color-map (red), in-lay. For significance see Figs. 3, 4, 5.