. 2015 Jun 1;10(6):e0128115. doi: 10.1371/journal.pone.0128115

Table 3. Comparison of the secure threshold $\overline{p}$ with the clustering coefficient-based threshold C* [67] and the statistical thresholds based on the adjusted p-values B0.01, B0.05 or B0.1 on a collection of synthetic and array datasets.

Dataset type	n	m	C*	B0.01	B0.05	B0.1	$\overline{p}$
Dataset type	n	m	C*	B0.01	B0.05	B0.1	$\overline{p}$	Simulated	50	1000	0.57	0.58	0.54	0.52	0.6152
Simulated	25	1000	0.69	0.76	0.72	0.70	0.7956
H-U133P	23	897	0.72	0.78	0.74	0.72	0.8125
H-U133P	10	897	0.78	0.96	0.94	0.93	0.9723
H-U133P	10	675	0.77	0.96	0.93	0.92	0.9681
H-U133P	9	897	0.79	0.97	0.96	0.95	0.9821
H-U133P	8	897	0.81	0.98	0.97	0.96	0.98999
H-U133P	7	897	0.81	0.99	0.99	0.98	0.99558
H-U133P	6	897	0.86	>0.99	>0.99	0.99	0.99872
H-U133P	5	897	0.92	>0.99	>0.99	>0.99	0.99984
H-U133P	4	897	0.99	>0.99	>0.99	>0.99	> 0.9999
H-U133A	4	675	0.99	>0.99	>0.99	>0.99	> 0.9999
H-I6	4	675	0.99	>0.99	>0.99	>0.99	> 0.9999
M-U74	4	401	0.97	>0.99	>0.99	>0.99	0.9999

Table 3. Comparison of the secure threshold p‾ with the clustering coefficient-based threshold C* [67] and the statistical thresholds based on the adjusted p-values B0.01, B0.05 or B0.1 on a collection of synthetic and array datasets.