HDNA methylation data-based molecular subtype classification related to the prognosis of patients with hepatocellular carcinoma

• DOI: 10.1186/s12920-020-00770-5
• Publication Date: 2020-08-24T12:03:37Z
• Authors: Hui He; Di Chen; Shimeng Cui; Gang Wu; Hailong Piao; Xun Wang; Peng Ye; Shi Jin
• Keywords: DNA methylation; Carcinoma, Hepatocellular; Hepatocellular carcinoma; Gene Expression Profiling; Liver Neoplasms; Bayes Theorem; QH426-470; DNA Methylation; TCGA database; RC31-1245; Epigenesis, Genetic; 3. Good health; Gene Expression Regulation, Neoplastic; 03 medical and health sciences; 0302 clinical medicine; Genetics; Biomarkers, Tumor; Humans; CpG Islands; Promoter Regions, Genetic; Internal medicine; Molecular subtype classification; Research Article

Abstract Background DNA methylation is a common chemical modification of DNA in the carcinogenesis of hepatocellular carcinoma (HCC).

Methods In this bioinformatics analysis, 348 liver cancer samples were collected from the Cancer Genome Atlas (TCGA) database to analyse specific DNA methylation sites that affect the prognosis of HCC patients.

Results 10,699 CpG sites (CpGs) that were significantly related to the prognosis of patients were clustered into 7 subgroups, and the samples of each subgroup were significantly different in various clinical pathological data. In addition, by calculating the level of methylation sites in each subgroup, 119 methylation sites (corresponding to 105 genes) were selected as specific methylation sites within the subgroups. Moreover, genes in the corresponding promoter regions in which the above specific methylation sites were located were subjected to signalling pathway enrichment analysis, and it was discovered that these genes were enriched in the biological pathways that were reported to be closely correlated with HCC. Additionally, the transcription factor enrichment analysis revealed that these genes were mainly enriched in the transcription factor KROX. A naive Bayesian classification model was used to construct a prognostic model for HCC, and the training and test data sets were used for independent verification and testing.

Conclusion This classification method can well reflect the heterogeneity of HCC samples and help to develop personalized treatment and accurately predict the prognosis of patients.