A novel defined N7-methylguanosine-related LncRNAs signature for predicting the prognosis of Hepatocellular carcinoma

Abstract

Background. Hepatocellular carcinoma (HCC) is a highly heterogeneous disease with poor prognosis, making the prediction of the prognosis much challenges. N7-methylguanosine (m7G) is a common modification with roles in eukaryotes. However, the value of m7G-related LncRNAs in predicting prognosis of HCC still remains to be further elucidated. This study aimed to find new potential biomarkers for predicting prognosis and precision medication. Methods. To build reliable predictive models,we obtained transcriptome and clinical data of HCC from The Cancer Genome Atlas (TCGA). M7G-related prognostic lncRNAs were selected by coexpression analysis and univariate Cox regression. The least absolute shrinkage and selection operator (LASSO) was utilized to construct the m7G related lncRNA model. Patients in TCGA cohort were divided into train group and test group. Applying the muti-lncRNA model,the patients were divided into high risk group and low risk group .The Kaplan–Meier analysis,time-dependent receiver operating characteristics (ROC), univariate Cox (uni-Cox) regression, multivariate Cox (multi-Cox) regression, nomogram, clinical characteristics and calibration curves were made to verify and evaluated the model. Furthermore ,we used principal component analysis (PCA), t-distributed stochastic-neighbor embedding (tSNE), immune analysis, and prediction of the half-maximal inhibitory concentration (IC50) to verify the signature in two risk groups .Results. A model with 4 m7G-related lncRNAs was constructed. The HCC patients were divided into two risk groups according to the model.Patients in the low-risk group showed significantly better outcome compared with patients in the high-risk group both in the train group and the test group (P < 0.001 in the TCGA train cohort and P = 0.001 in the test cohort). The risk score was an independent predictor for prognosis in multivariate Cox regression analyses (HR> 1, P< 0.01). Receiver operating characteristic (ROC) curve analysis show predictive capacity. Analysis of clinical characteristics and tumor microenvironment (TME) between risk groups showed significant difference (p=0.038). Functional analysis revealed different immune-related pathways were enriched, and immune status were different between two risk groups. Mediation analysis with IC50 revealed that the two risk group were significantly different, which could be a guidance of systemic treatment. Conclusion. A novel m7G-related LncRNAs signature can be used for prognostic prediction in HCC. Targeting m7G may be a therapeutic alternative for HCC.