Skip to main content

Table 2 The methylation status of the five CpG sites in the TCGA dataset and the validation dataset

From: Targeted bisulfite sequencing identified a panel of DNA methylation-based biomarkers for esophageal squamous cell carcinoma (ESCC)

 

CpG site

Gene

Position

Relation to CpG_Island

McaMb

McoMb

P valuec

log10(OR)d

95% CId

Sens

Spec

AUC

TCGA

cg15830431

STK3

chr8:99952591

Island

0.28

0.09

2.20E−04

4.11

1.91–7.43

0.65

0.94

0.82

cg19396867

NAa

chr19:40314862

N_Shore

0.45

0.20

3.60E−04

1.85

0.78–3.21

0.85

0.75

0.79

cg20655070

NAa

chr19:40315011

Island

0.44

0.19

1.71E−03

1.61

0.67–2.72

0.64

0.88

0.75

cg26671652

ZNF418

chr19:58446312

N_Shore

0.35

0.16

5.77E−04

1.95

0.67–3.61

0.86

0.75

0.78

cg27062795

ZNF542

chr19:56879613

Island

0.43

0.17

3.60E−04

2.93

1.65–4.44

0.86

0.81

0.80

Validation

cg15830431

STK3

chr8:99952591

Island

0.20

0.07

1.25E−06

3.04

1.82–4.53

0.66

0.77

0.71

cg19396867

NA

chr19:40314862

N_Shore

0.37

0.12

2.71E−11

2.83

1.93–3.91

0.65

0.88

0.80

cg20655070

NA

chr19:40315011

Island

0.31

0.09

8.04E−10

3.01

2.02–4.22

0.62

0.89

0.77

cg26671652

ZNF418

chr19:58446312

N_Shore

0.32

0.11

4.82E−11

3.20

2.18–4.39

0.58

0.93

0.79

cg27062795

ZNF542

chr19:56879613

Island

0.43

0.12

1.23E−12

2.55

1.77–3.50

0.72

0.82

0.83

  1. The sensitivity and specificity, as well as AUC, were both with a logistic regression prediction model without adjustment for gender, age, and smoking status and alcohol status
  2. Sens sensitivity, Spec specificity, AUC area under the curve
  3. aNA indicated that the CpG site is located outside of the coding region of the gene
  4. bMcaM represents the mean methylation percentage of the cases, and the McoM represents the mean methylation percentage of the controls
  5. c P value is calculated through the Wilcoxon rank-sum test followed by FDR (false discovery rate) adjustment for multiple correction
  6. dOR and 95% CI were determined by logistic regression