A novel coiled-coil domain containing-related gene signature for predicting prognosis and treatment effect of breast cancer.

Abstract

Breast cancer (BRCA) is a prevalent tumor worldwide. The association between the coiled-coil domain-containing (CCDC) protein family and different tumors has been established. However, the prognostic significance of this protein family in breast cancer remains uncertain. Gene expression and clinical data were obtained from the TCGA, METABRIC, and GEO databases. Prognosis genes were identified using univariate Cox and LASSO Cox regression, leading to the establishment of a prognostic signature. Subsequently, the risk model was conducted based on survival and clinical feature analyses, and a nomogram for prognosis prediction was developed. Furthermore, analyses of biological function, immune characteristics, and drug sensitivity were performed. Finally, single-cell sequencing data were utilized to uncover the expression patterns of genes in the risk model. Five genes were identified and utilized for risk modeling. The model demonstrated excellent prognostic value as indicated by ROC and Kaplan-Meier analysis. The high-risk group exhibited shorter survival time and higher likelihood of recurrence. Functional annotation indicated a correlation between the risk score and immune pathways. Conversely, the low-risk group displayed a greater enrichment in immune pathways and exhibited more active immune microenvironment characteristics. Additionally, drug sensitivity analysis using both public and our sequencing data revealed that the risk model possessed a broad range of predictive values. We have developed a gene signature and have verified that patients with low-risk are more likely to have better prognosis and respond positively to therapy. This finding offers a valuable point of reference for BRCA individualized treatment.