基于血糖相关基因不同亚型乳腺癌的预后模型

Prognostic Model for Breast Cancer Based on Blood Glucose-Related Genes in Different Subtypes

摘要: 乳腺癌的发病率在全球范围内呈上升趋势，也是女性中最常见的恶性肿瘤之一。本研究从TCGA数据库获取1028例乳腺癌患者的转录组数据以及相关的临床数据。在MSigDB数据库使用关键词筛选845个血糖相关的基因。对TCGA数据库中的肿瘤样本和正常样本进行差异表达分析获取7549个差异表达基因，同时将血糖相关基因集和差异表达基因取交集获得342个交集基因。使用加权基因共表达网络(WGCNA)筛选出与乳腺癌预后相关性高的黄色模块，并基于黄色模块中的基因使用一致性聚类算法对TCGA-BRCA乳腺癌样本进行最优化分组。通过单因素Cox回归分析和LASSO回归分析筛出10个与预后显著相关的基因(CLIC6、ELOVL2、KLHDC7B、KRT80、MMP12、NEK10、SEC14L2、SHCBP1、SUSD3和TAT)，并基于这10基因构建了乳腺癌的风险评分模型。本研究通过一致性聚类分组后构建的乳腺癌风险评分模型也能够较好地预测患者的预后情况，为乳腺癌的预后评估及开发可能的治疗靶点提供一定依据。

Abstract: The incidence of breast cancer is on the rise globally and remains one of the most common malignant tumors among women. In this study, transcriptomic data and relevant clinical data from 1028 breast cancer patients were obtained from the TCGA database. A total of 845 blood glucose-related genes were screened using keywords from the MSigDB database. Differential expression analysis between tumor and normal samples in the TCGA database yielded 7549 differentially expressed genes. The intersection of blood glucose-related genes and differentially expressed genes resulted in 342 overlapping genes. A weighted gene co-expression network was employed to identify the yellow module, which exhibited a high correlation with breast cancer prognosis. Based on the genes within the yellow module, an optimal grouping of TCGA-BRCA breast cancer samples was achieved using a consensus clustering algorithm. Univariate Cox regression analysis and LASSO regression analysis identified 10 genes significantly associated with prognosis (CLIC6, ELOVL2, KLHDC7B, KRT80, MMP12, NEK10, SEC14L2, SHCBP1, SUSD3, and TAT). A risk scoring model for breast cancer was subsequently constructed based on these 10 genes. The risk scoring model developed after consensus clustering grouping demonstrated a robust ability to predict patient prognosis, offering valuable insights for prognostic assessment and the potential development of therapeutic targets in breast cancer.