BMC Bioinformatics

Table 2 The estimated power of the five tests for p = 200, n = 80, and m = 50 based on 200 replications

From: A comparative study on gene-set analysis methods for assessing differential expression associated with the survival phenotype

			Case(A)					Case (B)
Case	c _p	m _p	GSEA1	GSEA2	GT	WT	GBST	GSEA1	GSEA2	GT	WT	GBST
(I)	0.0	0.1	0.145	0.185	0.295	0.275	0.250	0.120	0.185	0.380	0.380	0.345
		0.3	0.290	0.370	0.650	0.630	0.635	0.250	0.355	0.750	0.745	0.805
		0.5	0.400	0.450	0.825	0.805	0.820	0.400	0.525	0.900	0.905	0.950
	0.3	0.1	0.145	0.155	0.155	0.165	0.155	0..65	0.110	0.240	0.270	0.265
		0.3	0.245	0.225	0.430	0.420	0.415	0.205	0.230	0.510	0.520	0.540
		0.5	0.335	0.330	0.615	0.605	0.620	0.275	0.395	0.725	0.695	0.765
(II)	0.0	0.1	0.165	0.225	0.425	0.485	0.320	0.075	0.135	0.330	0.320	0.295
		0.3	0.890	0.965	1.000	1.000	0.980	0.235	0.375	0.755	0.745	0.755
		0.5	1.000	1.000	1.000	1.000	1.000	0.430	0.605	0.900	0.890	0.925
	0.3	0.1	0.140	0.190	0.310	0.330	0.215	0.100	0.080	0.270	0.285	0.255
		0.3	0.730	0.815	0.955	0.970	0.880	0.215	0.295	0.520	0.520	0.570
		0.5	0.985	0.995	1.000	1.000	1.000	0.265	0.395	0.670	0.640	0.745
(III)	0.0	0.1	0.120	0.115	0.310	0.335	0.245	0.125	0.135	0.270	0.310	0.280
		0.3	0.355	0.460	0.790	0.765	0.735	0.270	0.375	0.745	0.725	0.805
		0.5	0.610	0.615	0.895	0.885	0.885	0.385	0.495	0.885	0.890	0.925
	0.3	0.1	0.085	0.095	0.215	0.195	0.190	0.070	0.075	0.215	0.220	0.225
		0.3	0.260	0.330	0.610	0.585	0.535	0.165	0.275	0.550	0.505	0.595
		0.5	0.390	0.445	0.710	0.700	0.725	0.305	0.350	0.735	0.740	0.810
(IV)	0.0	0.1	0.105	0.205	0.440	0.460	0.300	0.100	0.170	0.385	0.415	0.360
		0.3	0.815	0.885	0.990	0.990	0.965	0.255	0.340	0.755	0.755	0.760
		0.5	0.995	1.000	1.000	1.000	1.000	0.440	0.550	0.875	0.890	0.930
	0.3	0.1	0.160	0.215	0.305	0.310	0.265	0.135	0.165	0.240	0.255	0.235
		0.3	0.585	0.760	0.925	0.920	0.830	0.185	0.285	0.560	0.550	0.555
		0.5	0.960	0.970	1.000	1.000	0.985	0.320	0.410	0.680	0.685	0.760

Note: m = the size of gene set; c_p= censoring proportion; m_p= the proportion of significant genes in a gene set. GSEA1 = GSEA ith the equal weight; GSEA2 = weighted GSEA; GT = Global Test; WT = Wald-type Test; GBST = Global Boost Test. For Case (I), all genes are independent; for Case (II), there is within-correlation among significant genes; for Case (III), there is an autoregressive correlation between significant genes; and for Case (IV), there is an unstructured correlation between significant genes

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com