Gene set analysis using variance component tests

Huang, Yen-Tsung; Lin, Xihong

doi:10.1186/1471-2105-14-210

BMC Bioinformatics

Table 1 The simulation results comparing size and power of TEGS and GSEA

From: Gene set analysis using variance component tests

	Alternative, p = 10				Null, p = 10						Alternative, p = 40				Null, p =40
		Correlated			Correlated			Correlated				Correlated			Correlated			Correlated
No. of β = -1.0		0	0	1	2	0	0	0	0	0	0	0	0	2	5	0	0	0	0	0	0
No. of β = 1.0		0	0	3	6	0	0	0	0	0	0	0	0	8	25	0	0	0	0	0	0
No. of β = -0.5		1	2	0	0	0	0	0	0	0	0	2	5	0	0	0	0	0	0	0	0
No. of β = 0.5		3	6	0	0	0	0	0	0	0	0	8	25	0	0	0	0	0	0	0	0
No. of β = 0		6	2	6	2	10	10	10	10	10	10	30	10	30	10	40	40	40	40	40	40
TEGS(True)	Permutation	21.2	32.8	70.3	87.7	4.8	5.3	5.5	5.3	5.3	5.3	27.6	65.8	96.4	98.4	5.4	4.1	4.9	5.6	4.5	5.7
TEGS(True)	$κ χ_{ν}^{2}$	22.0	33.9	70.9	88.0	4.8	5.7	5.5	5.9	5.3	5.6	27.7	67.8	96.6	99.4	5.9	4.2	4.9	5.9	4.8	5.6
TEGS(Indpt)	Permutation	11.0	20.0	41.5	75.0	6.0	6.4	5.1	6.4	5.8	5.5	9.3	23.5	34.8	91.7	3.9	6.6	4.9	3.4	6.1	5.6
TEGS(Indpt)	$κ χ_{ν}^{2}$	10.4	19.1	41.1	75.9	5.3	6.2	4.9	5.5	6.2	4.8	10.3	24.5	35.7	93.0	5.1	6.9	5.4	4.0	6.3	5.8
TEGS(Unstr)	Permutation	15.2	21.9	46.4	73.4	4.8	6.5	5.0	4.7	6.2	5.0	16.8	41.8	68.3	98.9	4.6	5.9	3.7	4.9	5.7	6.2
TEGS(Unstr)	$κ χ_{ν}^{2}$	13.0	20.8	45.1	73.0	4.3	6.1	4.6	4.4	5.7	4.9	17.5	42.4	70.5	98.9	5.5	6.1	5.0	5.2	6.4	6.4
TEGS(F-2)	Permutation	15.3	25.4	51.2	79.0	5.7	6.0	5.6	5.2	5.1	5.1	17.9	49.3	79.1	99.7	4.3	5.2	5.1	4.6	4.7	5.6
TEGS(F-2)	$κ χ_{ν}^{2}$	15.4	24.5	50.5	78.4	5.6	5.6	5.3	4.8	5.0	5.2	19.9	50.7	80.8	99.8	4.7	5.2	5.0	4.9	6.0	6.2
TEGS(F-adpt)	Permutation	14.5	24.1	48.3	75.9	4.6	7.0	5.1	5.3	6.4	5.3	18.2	45.8	73.7	99.5	4.9	5.9	4.3	5.2	5.5	6.3
TEGS(F-adpt)	$κ χ_{ν}^{2}$	13.7	22.8	48.0	75.1	5.0	6.5	4.5	5.1	5.9	5.0	18.9	46.1	74.7	99.6	4.9	5.5	5.4	5.6	5.2	6.5
TEGS(CpSym)	Permutation	14.5	24.6	53.1	84.1	5.2	5.6	5.6	5.3	5.8	5.9	16.5	34.5	71.1	99.1	4.2	4.5	4.4	3.1	5.0	5.1
TEGS(CpSym)	$κ χ_{ν}^{2}$	15.2	24.2	52.3	83.1	4.9	5.3	5.0	5.4	5.0	5.1	16.3	36.7	72.1	99.2	4.3	4.6	4.8	3.9	5.3	5.1
GSEA, no. of genes = 500		2.9	5.6	6.2	17.2	3.9	3.2	3.1	4.2	3.1	3.4	3.2	15.3	6.1	81.4	4.2	4.0	4.3	3.7	3.9	3.4
GSEA, no. of genes = 5,000		6.4	7.7	8.1	18.5	5.3	5.4	5.6	4.8	4.7	5.4	5.1	13.1	6.1	39.4	4.4	4.8	4.2	4.6	4.9	4.4

The entries are the percentage of times the p-values are less than 0.05. The columns with non-zero β’s represent the setting under the alternative and those with zero β’s represent the settings under the null. Abbreviations of the working covariance are the same as those in Figure 1. Two settings are considered: the first has 500 genes with 20 gene sets (8 gene sets under the alternative and 12 null gene sets of sizes 10 and 40); the second setting has 5000 genes with 20 gene sets (8 gene sets under the alternative and 12 null gene sets of sizes 10 and 40) and 4500 additional null genes. The power of TEGS for a given gene set is the same for the two settings.

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com