基于m5C相关基因构建三阴性乳腺癌预后预测模型及药物敏感性分析  

Prognosis Prediction Model and Drug Sensitivity Analysis of Triple-negative Breast Cancer Based on m5C Related Genes

作　　者：安外尔·约麦尔阿卜拉 布尔兰·叶尔肯别克 孙莉莉 刘富中 迪丽娜尔·叶尔夏提 郭文佳[1] ANWAIER Yuemaierabola;BUERLAN Yeerkenbieke;SUN Lili;LIU Fuzhong;DILINAER Yeerxiati;GUO Wenjia(Affiliated Cancer Hospital of Xinjiang Medical University,Urumqi 830011,China)

机构地区：[1]新疆医科大学附属肿瘤医院,乌鲁木齐830011

出　　处：《生物技术进展》2024年第1期149-159,共11页Current Biotechnology

基　　金：新疆维吾尔自治区自然科学基金杰出青年科学基金项目(2022D01E27);新疆维吾尔自治区天池英才项目(2022TCYCGWJ)。

摘　　要：为了探讨5-甲基胞嘧啶(5-methylcytosine,m5C)相关基因在三阴性乳腺癌(triple negative breast cancer,TNBC)患者治疗及预后中的潜在价值,构建了基于m5C相关基因的预后预测模型,用于评估TNBC患者的预后和生存状况。从基因表达总库(gene expression omnibus,GEO)数据库和癌症基因组图谱(the cancer genome atlas,TCGA)数据库中下载TNBC基因表达谱和相应的临床数据。通过Pearson分析确定了99个m5C相关基因,进一步采用单因素Cox分析鉴定出5个与预后有关的m5C相关基因(SLC6A14、BCL11A、UGT8、LMO4、PSAT1)并构建了风险评分(risk score)预测模型,根据风险评分中位值将患者划分为高风险组和低风险组。使用Kaplan-Meier(K-M)生存分析、受试者工作特征(receiver operating characteristic,ROC)曲线、多变量Cox回归分析、构建列线图和校准曲线评估了模型的预测效能。训练集和验证集的K-M生存曲线、受试者工作特征曲线下面积(area under the curve,AUC)分析均验证了模型具有良好的预测能力。多变量Cox回归分析显示,风险评分可作为独立的预后生物标志物。使用ssGSEA、免疫评分分析和化疗药物对高低风险组患者的半最大抑制浓度(half maximal inhibitory concentration,IC50)值差异分析显示,免疫细胞和免疫检查点基因以及大多数化疗药物的IC50值在不同风险组之间的表达存在显著差异。研究结果构建了基于5个m5C相关基因的风险评分预后预测模型,这将有助于阐明TNBC中m5C相关基因的作用机制,进而提供更有价值的预后及诊断的生物标志物和潜在的治疗靶点,为TNBC患者临床个性化治疗提供理论指导。In order to explore the potential value of 5-methylcytosine(m5C)-related genes in the prognosis of patients with triple negative breast cancer(TNBC)and in the treatment of TNBC,a prognostic prediction model based on m5C-related genes was constructed for assessing the prognosis and survival of TNBC patients.TNBC gene expression profiles and corre⁃sponding clinical data were downloaded from the gene expression omnibus(GEO)database and the cancer genome atlas(TCGA)database.99 m5C related genes were identified through pearson correlation analysis and one-way cox analysis was used to identify m5C-related genes,from which five prognosis-related m5C-related genes(SLC6A14,BCL11A,UGT8,LMO4,PSAT1)were identified and a risk score prediction model was constructed,and the patients were classified into high-risk and low-risk groups according to the median value of the risk score.The predictive efficacy of the model was assessed using Kaplan-Meier(K-M)survival analyses,receiver operating characteristic(ROC)curves,multivariate Cox re⁃gression analyses,construction of column line plots and calibration curves.The K-M survival curve,area under the curve(AUC)analysis of the training and validation sets verified the good predictive ability of the model.Multivariate Cox regression showed risk score could be used as an independent prognostic biomarker.Analysis of differences in half maximal inhibitory con⁃centration(IC50)values of chemotherapeutic agents using ssGSEA,immune score analysis and chemotherapeutic agents in pa⁃tients in the high and low risk groups showed significant differences in the expression of immune cells and immune checkpoint genes,as well as the IC50 values of most chemotherapeutic agents,between the different risk groups.In this study,we construct⁃ed a risk score prognostic prediction model based on five m5C-related genes,which would help to elucidate the mechanism of ac⁃tion of m5C-related genes in TNBC,and then provided more valuable biomarkers for prognosis and diagnosis and potential thera⁃peutic

关 键 词：m5C相关基因 三阴性乳腺癌 预后预测模型 化疗药物IC50 免疫浸润 免疫检查点 

分 类 号：Q75[生物学—分子生物学] R737.9[医药卫生—肿瘤]