Comparison of different cell type correction methods for genome-scale epigenetics studies

Kaushal, Akhilesh; Zhang, Hongmei; Karmaus, Wilfried J. J.; Ray, Meredith; Torres, Mylin A.; Smith, Alicia K.; Wang, Shu-Li

doi:10.1186/s12859-017-1611-2

BMC Bioinformatics

Table 3 Summary of sensitivity, specificity of Unadjusted, FaST-LMM-EWASher, RefFreeEWAS, SVA, ReFACTor and RefFreeCellMix for 100 simulated data across three settings

From: Comparison of different cell type correction methods for genome-scale epigenetics studies

	Sensitivity (Median, 95% interval)		Specificity (Median, 95% interval)
	Scenario 1 (ρ = 0)	Scenario 2 (ρ = 0.7)	Scenario 1 (ρ = 0)	Scenario 2 (ρ = 0.7)
	Number of Important CpGs =50
Unadjusted	0.960 (0.470, 1.000)	1.000 (1.000, 1.000)	1.000 (0.987, 1.000)	0.000 (0.000, 0.000)
Ewasher ^a	0.000 (0.000, 0.000)	0.000 (0.000, 0.000)	1.000 (0.999, 1.000)	1.000 (0.999, 1.000)
RefEWAS ^b	1.000 (0.960, 1.000)	0.000 (0.000,0.494)	0.997 (0.994, 0.999)	0.579 (0.055,1.000)
CellMix ^c	1.000 (0.980, 1.000)	1.000 (1.000, 1.000)	0.997 (0.993, 0.999)	0.546 (0.199, 0.923)
ReFACTor	1.000 (0.960, 1.000)	1.000 (1.000, 1.000)	0.996 (0.825, 1.000)	0.000 (0.000, 0.000)
SVA ^d	1.000 (0.980, 1.000)	1.000 (0.960, 1.000)	0.998 (0.996, 1.000)	0.998 (0.996, 1.000)
	Number of Important CpGs =100
Unadjusted	0.980 (0.664, 1.000)	1.000 (1.000, 1.000)	0.999 (0.976, 1.000)	0.000 (0.000, 0.000)
Ewasher ^a	0.000 (0.000, 0.000)	0.000 (0.000, 0.000)	1.000 (0.999, 1.000)	1.000 (0.999, 1.000)
RefEWAS ^b	1.000 (0.965,1.000)	0.000 (0.000,0.403)	0.995 (0.991, 0.998)	0.520 (0.014,1.000)
CellMix ^c	1.000 (0.975, 1.000)	1.000 (1.000, 1.000)	0.988 (0.968, 0.996)	0.211 (0.047, 0.525)
ReFACTor	1.000 (0.965, 1.000)	1.000 (1.000, 1.000)	0.994 (0.808, 0.998)	0.000 (0.000, 0.000)
SVA ^d	1.000 (0.990,1.000)	0.990 (0.965, 1.000)	0.996 (0.993, 0.999)	0.996 (0.993, 0.999)
	Number of Important CpGs =150
Unadjusted	0.993 (0.723, 1.000)	1.000 (1.000, 1.000)	0.999 (0.965, 1.000)	0.000 (0.000, 0.000)
Ewasher ^a	0.000 (0.000,0.000)	0.000 (0.000,0.000)	1.000 (0.999,1.000)	1.000 (0.999,1.000)
RefEWAS ^b	0.993 (0.973,1.000)	0.000 (0.000,0.294)	0.992 (0.986, 0.997)	0.496 (0.013,1.000)
CellMix ^c	1.000 (0.983, 1.000)	1.000 (1.000, 1.000)	0.975 (0.929, 0.993)	0.098 (0.022, 0.293)
ReFACTor	1.000 (0.980, 1.000)	1.000 (1.000, 1.000)	0.989 (0.794, 0.997)	0.000 (0.000, 0.000)
SVA ^d	1.000 (0.993,1.000)	0.993 (0.970, 1.000)	0.992 (0.988, 0.996)	0.993 (0.988, 0.996)

ρ = correlation between primary covariate and latent variables
ρ = 0 corresponds to data simulated from Scenario 1, while ρ = 0.7 corresponds to data simulated from Scenario 2
^aFasT-LMM-EWASher
^bRefFreeEWAS
^cRefFreeCellMix
^dSurrogate variable analysis

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com