基于可变剪切数据构建宫颈癌预后风险模型及验证

Building and validating a cervical cancer prognosis risk model based on alternative splicing data

目的 基于癌症基因组图谱(TCGA)数据库分析宫颈癌可变剪切(AS)事件及其相关基因,探讨其与患者预后生存的关系,构建AS事件预后风险模型并加以验证。方法 通过TCGASpliceSeq和TCGA数据库下载宫颈癌相关AS数据、RNA测序(RNA-Seq)数据和临床样本数据。通过R软件筛选并整理出94个有效样本进行分析,并对样本中7种类型AS事件进行研究。采用单因素Cox回归分析筛选与宫颈癌生存相关的AS事件;采用最小绝对收缩和选择算子(LASSO)回归分析构建基于AS的预后风险模型,并计算风险评分,将患者分为高风险组和低风险组;绘制生存曲线和受试者操作特征(ROC)曲线以评价模型效能,单/多因素Cox回归分析行预后独立分析;结合RNA-Seq数据,构建AS事件与剪切因子(SF)的调控网络。结果 在宫颈癌所有AS事件中,最常见的AS事件是外显子跳跃(ES)和可变启动子(AP)。在所有与宫颈癌预后相关的AS事件中ES类型最常见,外显子互斥(ME)类型最少,大部分AS事件均与生存显著关联。LASSO回归分析和Cox回归分析显示,FCF1和细胞质连接蛋白1(CLIP1)可变供体位点,蛋白酶C1抑制物(SERPING1)和视神经萎缩蛋白3(OPA3)可变启动子,NADH氧化还原酶辅酶3(NDUFA3)、非霍奇金淋巴瘤重复蛋白3(NHLRC3)、甘露糖苷酶2A类成员2(MAN2A2)外显子跳跃均是与宫颈癌患者预后高度相关的AS事件。Kaplan-Meier生存分析显示,高风险组患者预后生存显著低于低风险组患者生存。ROC曲线提示宫颈癌预后风险模型具有良好的预测性,AUC值均>0.9。预后模型的风险评分是宫颈癌的独立预后因素。AS事件与SF的调控网络表明,共有19个SF高度参与372个AS事件,其中多聚胞嘧啶结合蛋白(PCBP)、小核糖核蛋白35(SNRNP35)、环指蛋白40(RNF40)、小核糖核蛋白A(SNRPA)与多种AS事件相关。结论 AS事件与宫颈癌发生发展及预后关系密切,基于AS建立的预后风险模型能够有效预测宫颈癌患者的生存时间和生存状况,预测价值较高。与预后相关的AS基因及SF可作为宫颈癌治疗的潜在分子标志物。

Objective Based on the cancer genome atlas(TCGA)database,variable shear(AS)events and related genes of cervical cancer were analyzed,and the relationship between AS events and patients'prognosis and survival was dis-cussed,and the prognostic risk model of AS events was constructed and verified.Methods Download cervical cancer related AS data,RNA sequencing(RNA-Seq)data and clinical sample data from TCGA SpliceSeq and TCGA database.The R soft-ware was used to screen and sort out 94 valid samples for analysis,and 7 types of AS events in the samples were studied.Sin-gle-factor Cox regression analysis was used to screen for AS events related to cervical cancer survival.Least absolute shrink-age and selection operator(LASSO)regression analysis was used to construct an AS-based prognostic risk model,and the risk score was calculated to divide the patients into high risk group and low risk group.Survival curve and receiver operator char-acteristic(ROC)curve were drawn to evaluate the model effectiveness.Independent prognostic analysis was performed by Cox regression analysis with single and multiple factors.Combining RNA-Seq data,the regulatory network of AS events and shear factors(SF)was constructed.Results Among all AS events in cervical cancer,the most common AS events are exon skip(ES)and alternative promoter(AP).Among all the AS events associated with the prognosis of cervical cancer,ES type is the most common and mutually exclusive exons(ME)type is the least,and most AS events are significantly associated with sur-vival.LASSO regression analysis and Cox regression analysis showed that FCF1 and cytoplasmic linker protein 1(CLIP1)variable donor sites,plasma protease C1 inhibitor(SERPING1)and optic atrophin 3(OPA3)variable promoters,NADH dehydrogenase I alpha subcomplex 4(NDUFA3),NHL repeat-containing protein 3(NHLRC3),mannosidase alpha class 2A member 2(MAN2A2)exon hopping were all AS events highly correlated with the prognosis of cervical cancer patients.Kap-lan-Meier survival analysis showed that the survival of patients in the high-risk group was significantly lower than that in the low-risk group.ROC curve indicated that the prognostic risk model of cervical cancer had good prediction,and AUC values were all>0.9.The risk score of the prognostic model is an independent prognostic factor for cervical cancer.The regulatory network of AS events and SF showed that a total of 19 SFS were highly involved in 372 AS events,among which poly-cytosine binding proteins(PCBP),small nuclear ribonucleoproteins 35(SNRNP35),ring finger protein 40(RNF40)and small nuclear ribonucleoprotein A(SNRPA)were associated with multiple AS events.Conclusion AS events are closely related to the occurrence,development and prognosis of cervical cancer.The prognostic risk model established based on AS can effectively predict the survival time and survival status of cervical cancer patients,and has high predictive value.Prognostic AS gene and SF can be used as potential molecular markers for cervical cancer treatment.