Skip to main content

Table 4 Individual and joint discriminative performance of methylation at AHRR, 6p21.33, and F2RL3

From: Smoking-associated DNA methylation markers predict lung cancer incidence

Group

AUC (95% CI)

AHRR_cg05575921

6p21.33_cg06126421

F2RL3_cg03636183

Combinationa

Overall

Training set (n = 78 cases)

0.792 (0.736–0.848)

0.662 (0.597–0.726)

0.791 (0.735–0.846)

0.829 (0.778–0.881)

Validation set (n = 65 cases)

0.799 (0.733–0.866)

0.789 (0.725–0.853)

0.812 (0.725–0.871)

0.800 (0.737–0.861)

Age specific prediction

<65 years (n = 77 cases)

0.789 (0.728–0.850)

0.745 (0.687–0.803)

0.792 (0.735–0.849)

0.800 (0.745–0.856)

≥65 years (n = 66 cases)

0.790 (0.726–0.856)

0.677 (0.604–0.751)

0.793 (0.732–0.854)

0.817 (0.760–0.875)

Follow-up time-specific prediction

Initial 5 years (n = 68 cases)

0.791 (0.733–0.849)

0.696 (0.631–0.761)

0.808 (0.758–0.857)

0.812 (0.759–0.865)

Later years (n = 75 cases)

0.791 (0.734–0.849)

0.730 (0.673–0.786)

0.779 (0.722–0.837)

0.807 (0.755–0.859)

Histological subtype prediction

SCLC (n = 22 cases)

0.744 (0.630–0.858)

0.651 (0.535–0.767)

0.738 (0.632–0.843)

0.739 (0.634–0.844)

NSCLC (n = 119 cases)

0.802 (0.758–0.847)

0.721 (0.672–0.770)

0.798 (0.754–0.843)

0.823 (0.782–0.864)

Adenocarcinoma (n = 48 cases)

0.814 (0.751–0.877)

0.730 (0.659–0.800)

0.814 (0.751–0.876)

0.830 (0.770–0.891)

Squamous cell carcinoma (n = 38 cases)

0.787 (0.709–0.864)

0.731 (0.655–0.807)

0.769 (0.699–0.839)

0.786 (0.717–0.856)

Others (n = 32 cases)

0.775 (0.686–0.864)

0.673 (0.576–0.770)

0.800 (0.713–0.888)

0.813 (0.729–0.896)

  1. Abbreviations: AUC areas under the curve, CI confidence interval, SCLC small cell lung cancer, NSCLC non-small cell lung cancer
  2. aCombination formula: β1 × M AHRR  + β2 × M 6p21.33  + β3 × M F2RL3  + β4 × M 6p21.33  × β3 × M F2RL3  = (−0.0685) × cg05575921 + 0.4673 × cg06126421 + 0.3173 × cg03636183 + (−0.00612) × cg06126421 × cg03636183, where underlined coefficients were derived from regression coefficients in training set