NEAT: an efficient network enrichment analysis test

BMC Bioinformatics

Table 4 Results of simulation S5

Test	p ^KS	R ₁	R ₅	Sensitivity	Specificity	AUC
NEAT	0.343	0.62	0.98	79 %	95 %	0.925
NEA	0.024	0	0.82	73 %	96 %	0.912
LP	0	1.33	1.51	78 %	92 %	0.904
LA	0.111	1.16	1.33	73 %	93 %	0.908
LA+S	0.024	1.16	1.13	76 %	94 %	0.910
NP	0.323	1.42	1.16	70 %	94 %	0.908

The best results for each indicator are in bold. p ^KS denotes the p-value of the Kolmogorov-Smirnov test for uniform distribution, AUC is an abbreviation for “area under the ROC curve”. The distribution of p-values under H ₀ can be considered uniform for NEAT, LA and NP, and is questionable for LA+S. NEAT shows the highest values of sensitivity and AUC, and its specificity is exactly equal to the target value (95 %)

ISSN: 1471-2105