Table 1.

Classification of the network metrics tested. We separate the connectivity estimation methods by the inference process; whether or not only direct network edges are found (partial methods) or direct and indirect connections (marginal methods); whether or not directionality is ascribed to each edge; and whether or not the method is robust against spatial leakage artefacts. ^★PDC may be sensitive to magnetic field spread, see discussion in Methods Section 2.5.3.

Abbreviation	Connectivity metric	Type	Direct associations	Causal relations	Leakage-corrected
AEC	Amplitude envelope correlation	Amplitude coupling	Marginal	Undirected	Yes, with orthogonalisation
PAEC	Amplitude envelope partial correlation	Amplitude coupling	Partial	Undirected	Yes, with orthogonalisation
Coh	Absolute coherence	Spectral coherence	Marginal	Undirected	No
IMC	Imaginary coherency	Spectral coherence	Marginal	Undirected	Yes
PCoh	Partial coherence	Spectral coherence	Partial	Undirected	No
IMPC	Imaginary partial coherency	Spectral coherence	Partial	Undirected	Yes
PLV	Phase-locking value	Phase estimation	Marginal	Undirected	No
PLI	Phase lag index	Phase estimation	Marginal	Undirected	Yes
wPLI	Weighted phase lag index	Phase estimation	Marginal	Undirected	Yes
PSI	Phase slope index	Phase estimation	Marginal	Directed	Yes
MI	Mutual information between phases	Phase estimation	Marginal	Undirected	No
PDC	Partial directed coherence	Auto-regressive modelling	Partial	Directed	Yes★