Fig. 5: The contribution of between-system and within-system edges to effects of task factors and individual differences on subgraph 4 expression.
From: Functional brain network reconfiguration during learning in a dynamic environment
a The contribution of between-system and within-system edges to the effect of task factors on temporal relative expression of subgraph 4. To determine the relative contribution of between- and within-system edges on time-dependent subgraph 4 expression, we performed three comparisons on the effects estimated by different types of edges using t tests: within-system edges only (Within), between-system edges only (Between) and all edges (All). First, removing between-system edges (Within versus All) decreased the effect of CPP, RU and residual updating. Second, in contrast, after removing within-system edges (Between versus All), there was no significant change in these coefficients. Third, we directly compared the effects contributed from between-system edges only and from within-system edges only (Between versus Within). For between-system edges, there were stronger positive effects for CPP, RU, and residual updating. Error bars represent one SEM. (*p < 0.05, **p < 0.01, ***p < 0.001). b The contribution of between-system and within-system edges to the relationship between normative learning and dynamic modulation and average expression of subgraph 4. We performed the same three comparisons to determine the relative contribution of between- and within-system edges for each relationship with individual differences. For the effect of dynamic modulation, removing within-system edges (Between versus All) decreased the correlation coefficient. This correlation coefficient was also larger for within-system edges only than between-system edges only, but this effect was not statistically significant. For the effect of average expression, removing between-system edges (Within versus All) decreased the correlation coefficient, and the correlation coefficient was larger for between-system edges only than within-system edges only, though neither of these effects were statistically significant. Source data are provided as a Source Data file. Error bars represent one SEM. (**p < 0.01).
