Background:Tumor heterogeneity is closely related to the occurrence,progression and recurrence of renal clear cell carcinoma(ccRCC),making early diagnosis and effective treatment difficult.DNA methylation is an import...Background:Tumor heterogeneity is closely related to the occurrence,progression and recurrence of renal clear cell carcinoma(ccRCC),making early diagnosis and effective treatment difficult.DNA methylation is an important regulator of gene expression and can affect tumor heterogeneity.Methods:In this study,we investigated the prognostic value of subtypes based on DNA methylation status in 506 ccRCC samples with paired clinical data from the TCGA database.Differences in DNA methylation levels were associated with differences in T,N and M categories,age,stage and prognosis.Finally,the samples were divided into the training group and the testing group according to 450K and 27K.Univariate and multivariate Cox regression analysis was used to construct the prediction model in the training group,and the model was verified and evaluated in the testing group.Results:By univariate Cox regression analysis,21,122 methylation sites and 6,775 CpG sites were identified as potential DNA methylation biomarkers for overall survival of ccRCC patients(P<0.05).3,050 CpG sites independently associated with prognosis were identified with T,N,M,stage and age as covariables.Consensus cluster of 3,050 potential prognostic methylation sites was used to identify different DNA methylation subsets of ccRCC for prognostic purposes.We performed functional enrichment analysis on these 3,640 genes and identified 75 significantly enriched pathways(P<0.05).We then researched the expression of methylated genes in subgroups.Verifing with the training set,suggesting that DNA methylation levels generally reflect the expression of these genes.Conclusion:Based on TCGA database and a series of bioinformatics methods,We identified prognostic specific methylation sites and established prognostic prediction models for ccRCC patients.This model helps to identify novel biomarkers,precision drug targets and disease molecular subtypes in patients with ccRCC.Therefore,this model may be useful in predicting the prognosis,clinical diagnosis and management of patients with different epigenetic subtypes of ccRCC.展开更多
文摘Background:Tumor heterogeneity is closely related to the occurrence,progression and recurrence of renal clear cell carcinoma(ccRCC),making early diagnosis and effective treatment difficult.DNA methylation is an important regulator of gene expression and can affect tumor heterogeneity.Methods:In this study,we investigated the prognostic value of subtypes based on DNA methylation status in 506 ccRCC samples with paired clinical data from the TCGA database.Differences in DNA methylation levels were associated with differences in T,N and M categories,age,stage and prognosis.Finally,the samples were divided into the training group and the testing group according to 450K and 27K.Univariate and multivariate Cox regression analysis was used to construct the prediction model in the training group,and the model was verified and evaluated in the testing group.Results:By univariate Cox regression analysis,21,122 methylation sites and 6,775 CpG sites were identified as potential DNA methylation biomarkers for overall survival of ccRCC patients(P<0.05).3,050 CpG sites independently associated with prognosis were identified with T,N,M,stage and age as covariables.Consensus cluster of 3,050 potential prognostic methylation sites was used to identify different DNA methylation subsets of ccRCC for prognostic purposes.We performed functional enrichment analysis on these 3,640 genes and identified 75 significantly enriched pathways(P<0.05).We then researched the expression of methylated genes in subgroups.Verifing with the training set,suggesting that DNA methylation levels generally reflect the expression of these genes.Conclusion:Based on TCGA database and a series of bioinformatics methods,We identified prognostic specific methylation sites and established prognostic prediction models for ccRCC patients.This model helps to identify novel biomarkers,precision drug targets and disease molecular subtypes in patients with ccRCC.Therefore,this model may be useful in predicting the prognosis,clinical diagnosis and management of patients with different epigenetic subtypes of ccRCC.
