Construction of a prognostic model related to copper dependence in breast cancer by single-cell sequencing analysis

Abstract

Purpose: To explore the clinical significance of copper-dependent-related genes (CDRG) in female breast cancer (BC). Methods: CDRG were obtained by single-cell analysis of the GSE168410 dataset in the Gene Expression Omnibus (GEO) database. According to a 1:1 ratio, the Cancer Genome Atlas (TCGA) cohort was separated into a training and a test cohort randomly. Based on the training cohort, the prognostic model was built using COX and Lasso regression. The test cohort was used to validate the model. The GSE20685 dataset and GSE20711 dataset were used as two external validation cohorts to further validate the prognostic model. According to the median risk score, patients were classified as high-risk or low-risk. Survival analysis, immune microenvironment analysis, drug sensitivity analysis, and nomogram analysis were used to evaluate the clinical importance of this prognostic model. Results: 384 CDRG were obtained by single-cell analysis. According to the prognostic model, patients were classified as high-risk or low-risk in both cohorts. The high-risk group had a significantly worse prognosis. The area under the curve (AUC) of the model was around 0.7 in the four cohorts. The immunological microenvironment was examined for a possible link between risk score and immune cell infiltration. Veliparib, Selumetinib, Entinostat, and Palbociclib were found to be more sensitive medications for the high-risk group after drug sensitivity analysis. Conclusion: Our CDRG-based prognostic model can aid in the prediction of prognosis and treatment of BC patients.