Fuzzy Data Mining and Bioinformatics Analysis in Methylation Analysis of M6A Gene Promoter Region in Esophageal Cancer

This work was aimed at analyzing the correlation between methylation level of M6A gene in esophageal cancer (EC) and prognosis patients based on bioinformatics technology evaluating prognostic predictive values different data mining models. 80 EC healthy people were selected, serum collected to detect DNA methyltransferase. During radical resection EC, tumor tissues adjacent normal from gene. COX regression analysis employed analyze independent risk factors (IRFs) other treatments affecting patients. The particle swarm optimization (PSO) algorithm introduced improve fuzzy C -means clustering (FCM) algorithm. differences prediction efficiency logistic (LRA), decision tree (DT) C5.0, artificial neural network (ANN), support vector machine (SVM), improved FCM (IFCM) models compared. levels methyltransferase human histone deacetylase 1 (HSD-1) increased greatly ( id="M2"> P < 0.05 ). rates regulators (ALKBH5, HNRNPC, METTL3, WTAP, RBM15, YTHDC1, YTHDF1, FTO) obviously higher id="M3"> survival time high-risk much shorter than that low-risk id="M4"> Univariate multivariate showed gender, grade, TNM degree infiltration, ALKBH5, METTL3 genes IRFs for id="M5"> areas under ROC curve (AUCs) LRA, DT ANN, SVM, IFCM algorithms predicting 0.813, 0.857, 0.895, 0.926, 0.958, respectively, model had best diagnostic effect. In conclusion, detection no obvious patients, elevated an IRF addition, can be undertaken as preferred method