摘要
目的 基于铁死亡相关lncRNA构建结直肠癌预后模型.方法 以GSE39582和癌症基因组图谱(TCGA)数据库的结直肠癌患者,分别纳入训练队列和验证队列.从FerrDb网站下载铁死亡相关基因,并通过Pearson相关分析获得铁死亡相关lncRNA.采用单因素COX分析、套索算法(least absolute shrinkage and selection operator,LASSO)回归、log-rank检验和多元逐步COX回归分析筛选出预后关键lncRNA,构建风险评分公式.基于评分和临床特征,建立综合预后模型,使用Kaplan-Meier生存曲线、时间依赖性ROC曲线(time-dependent ROC curve,t-ROC)等在内部和外部队列中进行验证.使用ESTIMATE评分、单样本基因集富集分析算法、相关性分析等探索风险评分与免疫浸润和药物敏感性的相关性.结果 筛选出了9种lncRNA(NR2F1-AS1,SLC14A2-AS1,DNM1P35,KCNIP2-AS1,BVES-AS1,LINC00113,KCNQ1OT1,LINC00261,OGFRP1),构成风险评分.结合风险评分与3个临床特征(年龄、性别、M分期)的模型在训练队列中的1年、3年和5年ROC曲线下面积分别为0.81,0.80,0.78;在验证队列中分别为0.83、0.75和0.77.风险评分高的患者预后差,免疫细胞浸润、基质评分、PD-L2表达水平较高,对氟尿嘧啶的耐药性较低.结论 本研究构建的基于铁死亡相关lncRNA的预后模型可有效预测结直肠癌患者的预后.
Objective To establish a ferroptosis-related LncRNAs risk model to predict prognosis.Methods CRC patients in GSE39582 and TCGA were included into the training set and validation set,respectively.Ferroptosis genes were downloaded from FerrDb website,and ferroptosis-related IncRNAs were obtained using the Pearson correlation.The univariate COX analysis,the selection operator analysis(LASSO)regression,the log-rank test,and the stepwise COX regression analysis were utilized to screen the key ferroptosis-related IncRNAs and construct a prognostic signature.Then the integrated risk model based on the risk score and clinicopathological features was also established and would be validated in both internal and external cohorts using the Kaplan-Meier survival curves and the Time-dependent receiver operating characteristic curves.The ESTIMATE score,the single sample gene set enrichment analysis(ssGSEA),and the correlation analysis were performed to evaluate the association of risk score with the immune landscape and drug sensitivity.Results Nine prognostic lncRNAs(NR2F1-AS1,SLC14A2-AS1,DNMIP35,KCNIP2-AS1,BVES-AS1,LINC00113,KCNQ1OT1,LINC00261,OGFRP1)were filtered,constituting the prognostic signature.The risk model visualized by the nomogram was constructed of risk score and three clinicopathological features(age,gender,and Mstage),and its high accuracy was confirmed.In the train set,the AUCs corresponding to 1 year,3 years,and 5 years of survival were 0.81,0.80,and 0.78.And in the validation set,the AUCs corresponding to 1 year,3 years,and 5 years of survival were 0.81,080,and 0.78,respectively.Patients with higher risk scores showed higher levels of immune cell infiltration,stromal score,PD-L2 expression,and lower resistance to fluorouracil.Conclusion The prognostic risk model based on ferroptosis-related IncRNAs can effectively predict the prognosis of colorectal cancer patients.
出处
《浙江临床医学》
2023年第2期167-171,共5页
Zhejiang Clinical Medical Journal
基金
浙江省中医药科技计划项目(2021ZZ013)。
关键词
长非编码RNA
铁死亡
生存分析
结直肠癌
免疫浸润
Long non-coding RNA
Ferroptosis
Survival analysis
Colorectal cancer
Immuneinfiltration