Fig. 9.

Modeling the maximal similarity Fit(sFC, eFC) by the multiple linear regression (MLR) model with data variables from Fig. 7 as independent variables. (A1 - D1) Scatter plots with regression lines of the Fit-values predicted by MLR versus Fit(sFC, eFC) obtained by simulations of the phase model (1). The diagonals are depicted by thin black lines for comparison. (A2 - D2) The corresponding regression coefficients with the standard deviation for z-scored data obtained from the model fitting for parcellations (A) S200 and (B) HO96 0%, (C) Shen79 and (D) for joint data merged over all considered parcellations as indicated in the corresponding scatter plots. The gray bars indicate the regression coefficients, where the statistical significance with $p<.05$ was not achieved. The fractions of the explained variance $R^2$ are also shown in the scatter plots and in plot (E) for all individual parcellations for both phase and LC models as indicated in the legend. The dashed lines depict $R^2$ for the joint data also indicated in the legend.