Related Experiment Video
Updated: Oct 14, 2025
Genome-Wide Analysis of DNA Methylation in Gastrointestinal Cancer
Published on: September 18, 2020
Bioinformatics analysis to screen DNA methylation-driven genes for prognosis of patients with bladder cancer
Qing Zhou1, Qiuyan Chen2, Xi Chen1
1Central Laboratory, People's Hospital of Baoan District, The Second Affiliated Hospital of Shenzhen University, Shenzhen, China.
Background:
Bladder cancer (BLCA) is the most prevalent tumor affecting the urinary system, and has contributed to a rise in morbidity and mortality rates. Herein, we sought to identify the methylation-driven genes (MDGs)of BLCA in an effort to develop prognostic biomarkers suitable for the individualized assessment of patients with this particular cancer.
Methods:
The Cancer Genome Atlas (TCGA) dataset was distributed into training set (n=272) and testing set(n=117). The ConsensusClusterPluspackage was used to identify BLCA subtypes. The ChAMP package was used to analyze differential methylation probe (DMP) and differential methylation region (DMR). The differentially expressed genes (DEGs) were detected using DESeq2. Gene Ontology (GO) term enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were utilized to identify the pathways enriched of DEGs. Correlation analysis between 5'-C-phosphate-G-3's (CpGs) and DEGs was employed to identify the MDGs. The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) was used to build the protein-protein interaction (PPI) network of MDGs. Screening for BLCA prognosis-related MDGs and clinical features was conducted via the Cox regression model. A prognosis-related nomogram was developed and validated for prediction of the BLCA patients' survival.
Results:
We identified 2 BLCA clusters. Differential methylations at CpGs sites (dm-CpGs) were observed between cluster2 and cluster1, with 14,189 of them hypermethylated and 878 hypomethylated, predominantly in the CpG islands. In addition, a total 4,234 DEGs were identified between cluster2 and cluster1. The KEGG pathway and GO term enrichment analyses found that some DEGs were significantly enriched in multiple cancer-related pathways. A total of 33 MDGs were detected from correlation analysis between CpGs and DEGs. We selected BLCA-specific prognostic DMGs signatures for risk model development. The nomogram comprised a risk model to predict survival for BLCA patients. The efficiency of the prognostic prediction model was validated in the training and testing set.
Conclusions:
This study discovered differential methylation patterns and MDGs in BLCA patients, which provided a bioinformatics basis for guiding BLCA early diagnosis and prognosis analyses.
