An Eight-Gene-Methylation-Based Signature For Prognosis Of Bladder Cancer

Bladder cancer (BC) is one of the most common malignancies of the urological system with heterogeneous prognosis among patients. This study aimed to identify a novel prognostic model for BC using a joint analysis of DNA methylation and gene expression. A total of 412 cases with methylation data and 408 with mRNA expression data of BC patients were retrieved from the TCGA-BLCA Project. The data were preprocessed and the methylation data were randomly divided into a training set and a test set. Differentially methylated CpG sites were identified and a sure independence screening (SIS) model for BC survival prediction was built, and then was verified in the test set. The correlation of methylated CpG loci and associated gene expression was determined in the model. In addition, the relationship between expression levels of characteristic genes and overall survival rate was analyzed. As a result, the age, race and TNM staging were not significantly different between the training group and the test group. A total of 485 hypo-methylated sites and 808 hyper-methylated sites were identified. A SIS model containing 8 CpG sits was established and verified in the test set. Seven genes mapped by 8 CpG sits had a close relationship with overall survival of BC patients. Collectively, 8 CpG loci mapping to 7 genes were found to be a reliable and practical classifier to improve prediction accuracy of prognosis and survival in BC patients.