Construction and validation of a chromatin regulator-related lncRNA prognostic model for clear cell renal carcinoma
Objective To construct a prognostic model for chromatin regulator(CR)-related long non-coding RNA(lncRNA)in clear cell renal carcinoma(ccRCC),and improve the prognosis management of ccRCC.Methods Transcriptome data and clinical data of ccRCC were downloaded from TCGA database,and ccRCC samples were divided into training set(166 cases)and validation set(71 cases)at a ratio of 7∶3 using the"caret"R package.The differentially expressed CR and related lncR-NA were screened based on the"DESeq2"R package.A co-expression network between CR and lncRNA was constructed,and differentially expressed lncRNA with correlation coefficients|rs|＞0.3 and P＜0.05 were selected.Using univariate Cox,Lasso,and multivariate stepwise Cox regression analysis,a prognostic model for CR-related lncRNA was constructed.The risk scores of the training set and validation set were calculated,and ccRCC patients were divided into high-risk group and low-risk group based on the median risk score.The K-M curve and receiver operating characteristic(ROC)curve were used to evaluate the model,and univariate and multivariate Cox re-gression analyses were used to assess the independent prog-nostic performance of the risk score.An online prediction tool was developed using the"DynNom"and"shiny"R packages.Single-sample gene set enrichment analysis was used to ex-plore the relationship between the risk score and the immune microenvironment and immune checkpoints.Results In total,237 tumor samples and 72 adjacent normal tissue samples were included in this study.A total of 1 025 CR-related lncRNA were i dentified,and finally,seven CR-related lncRNA(DUXAP8,AC026462.3,LINC01460,AL592494.1,AL353804.2,AC012462.1,and AC009518.1)were selected to construct the prognostic model,which was developed into an online tool(https://xzlmodelshiny.shinyapps.io/DynNomapp/).The survival rate of the high-risk group was lower than that of the low-risk group(P＜0.05).The ROC curve showed that the prognostic performance of the model was good.The HR values of the risk score in the univariate analyse and multivariate analyse were 4.058(95%CI:2.530-6.508,P＜0.001)and 3.096(95%CI:1.887-5.080,P＜0.001),respectively.The proportion of immune-inhibitory cells such as MDSCs and Tregs in the high-risk group was higher than that in the low-risk group,and the enrichment levels of chemokines,immune checkpoints,and pro-inflammatory pathways were higher than those in the high-risk group(P＜0.05).In addition,the risk score was significantly positively correlated with immune checkpoint genes TNFRSF25 and TNFSF14(r＞0.3,P＜0.05).Conclusion The CR-relat-ed lncRNA risk model constructed in this study can effectively predict the prognosis of ccRCC patients independently.
NETL
NSTL
万方数据
