Fig. 1

Calculation of anti-symmetric functional networks. a As an example, we show the time activation series of 5 brain regions represented as nodes in the brain. b First, we calculate delayed correlation networks by computing the delayed Pearson’s correlation coefficient between all pairs of regions at varying delays; here, the network and the delayed connectivity matrix are shown at a delay of 1. Then, we split this delayed matrix into its c symmetric and d anti-symmetric component matrices. In our subsequent analyses, we use the anti-symmetric matrices as a representation of the whole-brain-directed functional connectivity. For all matrices and networks, darker colors and thicker lines represent stronger connections