Fig. 3: Patterns of connectivity in subgraphs.
From: Functional brain network reconfiguration during learning in a dynamic environment
a Converting edges between nodes into edges between systems. First, the edges of each subgraph can be folded into a square matrix, representing the edges between every pair of nodes (ROIs). Then, based on the 13 putative functional systems reported by Power et al. (2011), we categorized each edge according to the system(s) to which the two nodes (ROIs) belonged. We calculated the mean strength of edges linking a node in one system to a node in another system, and refer to that value as the between-system edge. Similarly, we calculated the mean strength of edges linking two nodes that both belong to the same system and refer to that value as the within-system edge. Edges between nodes and edges between systems were normalized into the scale between 0 and 1. b Edges between systems in the ten subgraphs identified by NMF. We show only significant edges (p < 0.05 after the Bonferroni correction for multiple comparisons). For each subgraph, the top matrix shows the significant edges in that subgraph within or between systems. For example, Subgraph 1 has high edge strengths along the diagonal; thus, this subgraph describes functional connectivity that lies predominantly within functional systems. In contrast, subgraph 5 has high edge strengths along a single row and column, corresponding to the visual system; thus, this subgraph describes functional connectivity between the visual system and all other systems. Subgraphs varied in the degree to which they represent interactions within the same system (e.g., subgraph 1) versus interactions between different systems (e.g., subgraph 10). All nodes from systems involved in significant edges are shown on the brain below by the BrainNet Viewer42 under the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
