Construction of a DNA damage repair gene signature for predicting prognosis and immune response in breast cancer

Abstract

DNA damage repair (DDR) genes are involved in developing breast cancer. Recently, a targeted therapeutic strategy through DNA repair machinery, including PARPi, has initially shown broad development and application prospects in breast cancer therapy. However, few studies that focused on the correlation between the expression level of DNA repair genes, prognosis, and immune response in breast cancer patients have been recently conducted. Herein, we focused on identifying differentially expressed DNA repair genes (DEGs) in breast cancer specimens and normal samples using the Wilcoxon rank-sum test. Biofunction enrichment analysis was performed with DEGs using the R software “cluster Profiler” package. DNA repair genes were involved in multivariate and univariate Cox regression analyses. After the optimization by AIC value, 11 DNA repair genes were sorted as prognostic DNA repair genes for breast cancer patients to calculate risk scores. Simultaneously, a nomogram was used to represent the prognostic model, which was validated using a calibration curve and C-index. Single-sample gene set enrichment analysis (ssGSEA), CIBERSORT algorithms, and ESTIMATE scores were applied to evaluate the immune filtration of tumor samples. Subsequently, anticarcinogen sensitivity analysis was performed using the R software “pRRophetic” package. Unsupervised clustering was used to excavate the correlation between the expression level of prognostic-significant DNA repair genes and clinical features. In summary, 56 DEGs were sorted, and their potential enriched biofunction pathways were revealed. In total, 11 DNA repair genes (UBE2A, RBBP8, RAD50, FAAP20, RPA3, ENDOV, DDB2, UBE2V2, MRE11, RRM2B, and PARP3) were preserved as prognostic genes to estimate risk score, which was applied to establish the prognostic model and stratified breast cancer patients into two groups with high or low risk. The calibration curve and C-index indicated that they reliably predicted the survival of breast cancer patients. Immune filtration analysis, anticarcinogen sensitivity analysis, and unsupervised clustering were applied to reveal the character of DNA repair genes between low- and high-risk groups. We identified 11 prognosis-significant DNA repair genes to establish prediction models and immune responses in breast cancer patients.