Table 2.

Experiment-wide Enrichment Results for Salience-Visual Functional Connectivity

	All Subjects			Excluding n = 8 sleeping subjects
	ORA	Max-Mean	GSEA	ORA	Max-Mean	GSEA
SA	27.9%	1.30	0.57	26.6%	1.29	0.56
p-value	0.00016*	0.00051*	0.00056*	0.00022*	0.00058*	0.000840
RRB	25.0%	1.33	0.56	20.0%	1.26	0.53
p-value	0.00022*	0.00037*	0.00072	0.0015	0.000947	0.002520
SA, RRB jointly	27.2%	N/A	0.65	27.2%	N/A	0.63
p-value	0.00001*		0.00020*	0.0001*		0.000360*

Observed values for the ORA (5% threshold), max-mean and GSEA enrichment statistics with p-values in italics: raw permutation p-value are presented, those which meet our p ~ 0.0005 threshold for significance are marked with an asterisk (*). Results are given for the entire sample and excluding sleeping subjects (n = 8). The p-values were determined using independent permutation runs: 25 K permutations for the more compute-intensive GSEA, and 250 K for ORA and max-mean each. Two-sided t-statistics were used to screen for association with SA or RRB individually. Partial F tests were used to screening for association with SA and RRB jointly (d.f. = 2, 85 for the complete sample and d.f. = 2, 77 when excluding sleeping subjects). The expectation for ORA is 5% under the null. Max-Mean is a directional test and cannot be used with F screening statistics. The median max-mean value under the null hypothesis of no enrichment is ~ 0.54 with 0.1% and 99.9% quantile estimates of 0.31 and 1.25 (obtained in separate analyses). The median GSA value under the null is ~ 0.21 (with 0.1% and 99.9% quantiles: 0.17 and 0.60) for the F test and ~ 0.20 (0.18–0.54) for t-tests (separate analysis as well)