Tu1769 – Dna Methylation Signatures in Crohn's Disease-Related Genes are Useful to Confirm Diagnosis and Activity

Receiver operating characteristic curve analysis the performance of multivariate Logistic regression model on training datasets and testing datasets.(a) AUROC of the genetic prediction model that fitted in training dataset is 0.794 (95% CI: 0.682-0.905,P=6.00×10-6), the sensitivity and specify is 81.5% and 58.8%, respectively, Figure 2 a.The genetic model was also verified in representative testing dataset, AUROC=0.812(95% CI: 0.714-0.910,P= 7.90×10-5), Figure 2 b.The combined genetic-clinical prediction model's AUROC in representative training dataset is 0.818 (95% CI: 0.716-0.921,P=9.36×10-7), the sensitivity and specify is 86.9% and 72.0%, respectively, Figure 2 c.The combined genetic-clinical prediction model's AUROC in representative testing dataset is 0.888 (95% CI: 0.812-0.963,P=9.52×10-7), Figure 2 d.A score nomogram based on entire dataset was created which can be easily and simply used to estimate Primary response of 14-week, it was developed with hs-CRP at second week, rs61740234, rs61886887, rs111504845, rs357291 and rs2269330, Figure 3. Tu1768