基于机器学习的肺腺癌细胞焦亡预后模型构建与验证  

Construction and validation for a prognostic model of pyroptosis from lung adenocarcinoma based on machine learning

作　　者：陈少明 胡艳 洪旭东 郑伟 胡旭钢 CHEN Shaoming;HU Yan;HONG Xudong;ZHENG Wei;HU Xugang(Department of General Practice,No.903 Hospital of PLA Joint Logistic Support Force(Affiliated Xihu Hospital Hangzhou Medical College),Hangzhou,Zhejiang 310000,China)

机构地区：[1]联勤保障部队第九〇三医院(杭州医学院附属西湖医院)全科医学科,浙江杭州310000

出　　处：《中华全科医学》2025年第1期21-25,共5页Chinese Journal of General Practice

基　　金：浙江省医药卫生科技计划项目(2021KY946);浙江省教育厅一般科研项目(Y202456549);杭州市医药卫生科技计划项目(B20252302)。

摘　　要：目的基于细胞焦亡相关基因(PRGs),构建肺腺癌预后模型,并评估其与免疫调节的关联。方法从TCGA数据库中获取肺腺癌基因表达和临床数据,通过DESeq2法获得差异表达PRGs;将患者分层抽样后按7∶3的比例划分为训练集和验证集,通过Cox回归和Lasso回归筛选目标因子并构建预后模型;在训练集、验证集和3个测试集(GSE30219、GSE31210和GSE50081)中通过生存分析、ROC曲线、Cox回归分析等评估预后模型的预测能力;绘制列线图观察预后模型联合临床特征在预后评估中的作用;分析预后模型与免疫细胞浸润的相关性。结果本研究构建了由CPA3、FAT1、MST1和TFAP2A等4个PRGs组成的肺腺癌预后模型。根据预后模型计算患者的风险评分,以中位值为界将患者分为高风险组和低风险组。高风险组患者预后较差(P<0.05);预后模型风险评分对肺腺癌患者具有良好的生存预测价值,并可作为独立的不良预后因素;基于预后模型和临床特征建立的列线图可有效预测患者的生存状态。免疫浸润分析表明高风险组患者处于免疫失调状态,CD274、CD276等免疫检查点分子表达显著升高。结论基于CPA3、FAT1、MST1和TFAP2A的预后模型可有效预测肺腺癌患者的预后,并与患者的免疫状态相关,可作为患者治疗及预后评价的指标。Objective To construct a prognostic model for lung adenocarcinoma(LUAD)based on pyroptosis related genes(PRGs)and evaluate its association with immune regulation.Methods LUAD gene expression and clinicopathological data were obtained from the TCGA database,and differentially expressed PRGs were obtained by the DESeq2 method;the patients were divided into training and validation sets in a 7∶3 ratio by stratified random sampling,and prognosisVrelated genes were screened,and prognostic models were constructed by Cox and Lasso regression;the predictive ability of the prognostic model was evaluated by Kaplan-Meier analysis,ROC curve,and Cox regression analysis in the training set,validation set,and the three test sets(GSE30219,GSE31210,and GSE50081);a nomogram model was constructed to observe the role of the prognostic model combined with the clinical features in the prognostic assessment of LUAD;the correlation between the prognostic model and the immune infiltration was analyzed.Results A prognostic model for LUAD consisting of four PRGs,including CPA3,FAT1,MST1,and TFAP2A,was constructed.Patients were divided into the high-risk group and the low-risk group based on median risk scores.Patients in the high-risk group had a poorer prognosis(P<0.05);The prognostic model risk score has good survival prediction value for patients and was an independent poor prognostic indicator;the nomogram model could effectively predict the survival status.Immune infiltration analysis showed that patients in the high-risk group were in a state of immune dysregulation,and the expression of CD274 and CD276 was significantly elevated.Conclusion The prognostic model based on CPA3,FAT1,MST1,and TFAP2A can effectively predict the prognosis of patients with LUAD and correlate with the immune status of the patients,which can be used as an indicator for the evaluation of patient treatment and prognosis.

关 键 词：肺腺癌 细胞焦亡 预后模型 列线图 免疫浸润 

分 类 号：R734.2[医药卫生—肿瘤] R730.7[医药卫生—临床医学]