构建与细胞焦亡和坏死性凋亡相关预后模型预测肝癌预后

目的 构建与焦亡和坏死性凋亡相关用于预测肝癌预后的模型。方法 通过癌症基因组图谱（The Cancer Genome Atlas，TCGA）获得的肝癌和非肝癌样本RNA测序数据以及相应的临床信息。通过GeneCards数据库检索得到关于焦亡和坏死性凋亡相关基因，R包“limma”筛选差异基因，R包“ConsensusClusterPlus”对肝癌进行聚类分型。之后再筛选不同肝癌分型之间的差异基因。基于差异基因，利用LASSO Cox回归筛选基因来构建预后模型。根据中位风险评分将肝癌患者分成高低风险组，两组间进行主成分（Principal Component Analysis，PCA），t分布随机邻域嵌入（t-Distributed Stochastic Neighbor Embedding，t-SNE），总生存时间（Overall Survival，OS）和受试者工作特征曲线（Receiver Operating Characteristic，ROC）分析。结果 通过一致性聚类分析将肝癌患者分为C1，C2两种分型，分型的OS高于C2分型的OS且有统计学差异（P<0.001）。筛选两分型间有关焦亡和坏死性凋亡相关差异基因，构建预后模型将肝癌患者分为低、高风险组，两组间OS有显著差异（P<0.001）。对风险评分进行独立预后分析，结果表明风险评分可作为独立因素预测肝癌预后，风险评分越高，肝癌患者预后越差（单因素HR=4.846，95%CI:2.950~7.971；多因素HR=4.227，95%CI:2.499~7.149。结论 本研究确定了与焦亡和坏死性凋亡相关的，可以独立预测肝癌患者的预后模型，包含九个基因（CASP8、TREM2、SQSTM1、ADORA1、ADORA2B、PARP1、MKI67、GLMN和POP1）。

Modles related to the pyroptosis and necroptosis for predicting the prognosis of liver cancer

Objective To Construct models associated with pyroptosis and necrotizing apoptosis for predicting the prognosis of liver cancer. Methods RNA sequencing data of liver cancer and non-liver cancer samples were obtained by The Cancer Genome Atlas (TCGA). The GeneCards database was searched for genes related to pyroptosis and necroptosis. The R package "limma" was used to screen for differential genes and the R package "ConsensusClusterPlus" was used to cluster and type hepatocellular carcinomas. Prognostic models were constructed by using LASSO Cox regression based on differential genes between different liver cancer types. Patients with liver cancer were divided into high and low risk groups based on median risk scores, and principal component (PCA), t-SNE, overall survival time (OS) and receiver operating characteristic (ROC) analyses were performed between the two groups. Results Liver cancer patients were classified into two subtypes (C1, C2) by consistent clustering analysis, and the OS of C1 type was higher than that of C2 type and there was a statistical difference (P<0.001). Genes related to differences in pyroptosis and necroptosis between the two types were screened, and a prognostic model was constructed to classify hepatocellular carcinoma patients into low-risk and high-risk groups, with a significant difference in OS between the two groups (P<0.001). Independent prognostic analysis of risk score showed that risk score can be used as an independent factor to predict liver cancer prognosis, and the higher the risk score is, the worse the prognosis of liver cancer will be (univariate analysis HR=4.846, 95%CI:2.950-7.971 and multivariate analysis HR=4.227, 95%CI:2.499-7.149). Conclusion This study performed a comprehensive bioinformatics analysis and identified a prognostic differentially expressed pyroptosis and necroptosis related genes signature containing nine genes (CASP8, TREM2, SQSTM1, ADORA1, ADORA2B, PARP1, MKI67, GLMN, and POP1) for liver cancer patients, which may play an important role in the progression of liver cancer.