Table 2 Prediction accuracy (standard errors) from the average of cross-validation AUC using all methylation data in different methods: Bayes-A, Bayes-B, Bayes-C, Bayesian LASSO (BL), and Bayesian ridge regression (BRR).

From: Integration of single nucleotide variants and whole-genome DNA methylation profiles for classification of rheumatoid arthritis cases from controls

Model

Testing population

Nonadjusted

Adjusted-cell1

Adjusted-all2

Bayes-A

0.868 (0.044)a

0.821 (0.050)b

0.657 (0.057)c

Bayes-B

0.868 (0.043)a

0.821 (0.051)b

0.659 (0.057)c

Bayes-C

0.867 (0.044)a

0.820 (0.050)b

0.660 (0.058)c

BL

0.865 (0.044)a

0.821 (0.048)b

0.662 (0.058)c

BRR

0.867 (0.044)a

0.820 (0.050)b

0.658 (0.057)c

  1. 1Methylation signatures adjusted for cell proportions only.
  2. 2Methylation signatures were adjusted for all available confounders including cell proportions, age, sex, and smoking status. Average AUCs in testing populations with different superscript(s) are significantly different (p value < 0.05).
Back to article page