Table 2. Classification accuracy for various neuroimaging markers (50% of the data are used as training set and the rest 50% as test set, except for the last column, where a leave-one-out cross-validation is used).

Neuroimaging markers	A: Community matrix K(400 edges)	B: Community Matrix K(50 edges)	C: Asymmetry d: 45-dimensional	D: Asymmetryρ: 45-dimensional	ECross correlation matrix(400 edges)	B+D	A+D	A+D(Leaveoneout)
Predictionaccuracy(mean±std)	77.6% ±3.47%	73.2% ±3.89%	75.8%± 3.64%	75.5% ±3.62%	70.5%± 4.48%	76.9%± 3.59%	80.2% ±3.45%	83.9%
Sensitivity(mean)	77.1%	72.3%	75.2%	74.8%%	69.7%	75.4%	78.4%	82.5%
Specificity(mean)	78.0%	73.7%	76.3%	76.1%	71.1%	78.1%	81.6%	85%

The best results are achieved when we combine the features from the community matrix K and the asymmetry measure ρ. The accuracy of classification using SVM versus the number of edges selected from the community matrix K can be found in the Text S4.