Construction and validation of a prognostic risk model for cuproptosis-related lncrna in breast cancer

Abstract

Abstract Background Long non-coding RNAs(LncRNAs) are vital in regulating programmed death in breast cancer. Cuproptosis is a newly type of regulatory cell death(RCD) in tumor pathogenesis, development, prognosis and potentially as a target for immunotherapy. Methods The expression profile data and clinical information data of breast cancer (BRCA) from the Cancer Genome Atlas (TCGA) were downloaded using the R package TCGA biolinks. The differential expression analysis was performed using R package ‘limma’. Univariate Cox analysis was performed on deregulated LncRNA related to cuproptosis to screen signature related to the prognosis of breast cancer. Using Cytoscape construct mRNA (cuproptosis-related factors) and co-expression of LncRNA networks. The network was verified in TCGA training set, global set and test set. Functional enrichment analysis was also performed for studied pathway. Results In this study, 1156 samples was screened from TCGA dataset for differential expression analysis, and finally a total of 284 dysregulated LncRNAs in breast cancer was identified. Then based on the expression of cuproptosis factors and dysregulated LncRNAs, a total of 140 cuproptosis-related dysregulated lncRNAs (CRDLs) were obtained by correlation analysis. Finally, eight model genes(LINC01235, MIR205HG, RP11-459E5.1, RP11-817J15.3, KLHDC7B-DT, LMNTD2-AS1, RP11-6O2.3, LINC00987) were obtained, one of which was a risk factor., we identified a prognostic risk model of dysregulated lncRNAs associated with cuproptosis by collecting RNA-seq data from TCGA. Subsequently, we verify this model, and the results showed that there were significant differences between the high and low-risk groups. Then, we took the test set and whole set, respectively, to perform model verification and prove that they have stable and favorable performance on survival prediction. And the Risk score we selected can be used as an independent prognostic factor. There were significant differences in the proportion of immune cell infiltration, genomic mutations, pathway enrichment scores, expression levels of immune checkpoints and chemoresistance between high and low-risk groups. Conclusion Ultimately, we conclude that the risk score can predict the benefit of immunotherapy in patients, and the model genes may be markers of immunotherapy response.