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

BMC Bioinformatics

Table 1 Number of significant CpG sites with and without cell type correction and overlap with the SVA method (data on prenatal arsenic exposure and DNA methylation)

Method	Identified CpGs (N)^#	Overlap with SVA (%)	p-value^##
Houseman et al.	10	1.20	<0.0001
minfi	57	4.62	<0.0001
SVA	498	–	–
RefFreeEWAS	133	6.01	<0.0001
RefFreeCellMix	2932	0.60	1.0
ReFACTor	58,871	13.03	1.0
EWASher ^a	0	0.0	–
RUV	356	0.20	1.0
Unadjusted ^b	3	0.60	<0.0001

^#The selection of CpG sites is based on FDR-adjusted p-values (FDR is controlled at 0.05)
^## P-value is based on Fishers exact test for overlap with results from SVA. The null hypothesis is that there is no overlap with the CpGs identified based on SVA
^aThe FasT-LMM-EWASher method
^bUnadjusted: cell type compositions were not included in the analyses

ISSN: 1471-2105