Skip to main content

Table 5 Performance metrics under simulation scenario 3 (opposite effects)

From: An evaluation of methods correcting for cell-type heterogeneity in DNA methylation studies

Method

Number of false positives

Power

KS

GIF

\(\hat {K}\)

Unadjusted

5

0.490

0.041

0.97

–

Ref-based

1

0.506

0.036

0.87

–

Ref-free

176

0.608

0.053

1.19

14

SVA

37

0.562

0.003

1.00

10

ISVA

49

0.558

0.006

1.01

14

EWASher

0

0.128

0.092

0.74

–

CellCDec

3

0.502

0.033

0.88

–

Deconf

2

0.522

0.035

0.89

–

RUV

23

0.532

0.012

1.00

37

  1. KS Kolmogorov–Smirnov statistic, GIF genomic inflation factor