乳腺癌代谢相关基因识别与预后模型构建

Identification of metabolism related genes and construction of prognosis model in breast cancer

目的探讨代谢相关基因表达在乳腺癌发生发展中的作用及预后价值。方法从京都基因和基因组百科全书(KEGG)中获取代谢相关基因。EdgeR软件包用于鉴定癌症基因组图谱(TCGA)中乳腺癌的差异表达基因(DEGs)。通过单变量Cox和Lasso-penalized Cox回归分析建立预后模型。通过基因表达综合数据库(GEO)建立有效的预测模型。利用列线图和受试者操作特征(ROC)曲线验证模型的准确性。定量实时PCR检测乳腺癌组织中关键基因表达。结果在乳腺癌中鉴定出与代谢相关的178个DEGs,并构建了一个包含14个基因的预后模型。在TCGA预后模型和GEO验证模型中将患者分为高风险组和低风险组,结果显示高风险组预后较差,进一步验证预测模型具有较高的准确性。乳腺癌组织中TSTA3、MTHFD2、P4HA1和SEPHS2表达上调(均P<0.001),INPP1表达下调(P=0.026)。结论本研究构建了一个与乳腺癌代谢基因相关的预后评估模型并验证了其可行性和准确性。该模型有助于更精准地评估不同患者对于代谢相关治疗的反应性,预测抗肿瘤疗效,为乳腺癌患者个体化治疗提供参考。

Objective To investigate the role and prognostic value of metabolism related gene expression in the occurrence and development of breast cancer.Methods Metabolism related genes were obtained from the Kyoto Encyclopedia of Genes and Genomes(KEGG).The EdgeR software package was used to identify differentially expressed genes(DEGs)in breast cancer in the Cancer Genome Atlas(TCGA).The prognostic model was established by univariate Cox and Lasso-penalized Cox regression analysis.An effective prediction model was established through the Gene Expression Omnibus(GEO)comprehensive database.The accuracy of the model was verified by column charts and receiver operating characteristics(ROC)curves.Quantitative Real time PCR(qRT PCR)was used to detect the expression of key genes in breast cancer tissues.Results A total of 178 DEGs related to metabolism were identified in breast cancer,and a prognostic model containing 14 genes was constructed.In the TCGA prognosis model and GEO validation model,patients were divided into high-risk and low-risk groups,and the results showed that the high-risk group had poor prognosis.Further validation of the prediction model showed high accuracy.The expression of TSTA3,MTHFD2,P4HA1 and SEPHS2 in breast cancer tissues was up-regulated(all P<0.001),and the expression of INPP1 was down-regulated(P=0.026).Conclusions This study constructed a prognostic evaluation model related to breast cancer metabolic genes and verified its feasibility and accuracy.This model can help to more precisely assess the responsiveness of different patients to metabolism-related treatments,predict anti-tumor efficacy,and provide reference for the personalized treatment of breast cancer patients.