基于生信分析和网络药理学的片仔癀治疗肝细胞癌的炎症相关分子靶点的鉴定和预后模型构建

Identification of Inflammation-related Molecular Targets and Construction of Prognostic Models for Pien Tze Huang in the Treatment of Hepatocellular Carcinoma Based on Biometric Analysis and Network Pharmacology

关键词:  片仔癀  肝细胞癌  炎症  预后模型  生信分析

Key words:  Pien Tze Huang  hepatocellular carcinoma  inflammation  prognostic model  biometric analysis

摘要:

目的 分析片仔癀治疗肝细胞癌的炎症相关分子靶点，对其机制进行初步探索。方法 通过TCMSP和BATMAN数据库获取片仔癀的成分和靶点。通过Genecards、OMIM和TCGA数据库获取肝细胞癌的疾病靶点。对化合物靶点和疾病靶点取交集，得到片仔癀治疗肝细胞癌的靶点。从GSEA数据库获取炎症通路的相关基因，然后对片仔癀治疗肝细胞癌的靶点和炎症相关基因进行相关性分析，筛选出炎症相关靶点，通过GO和KEGG富集分析探索其机制。然后进行单因素cox分析和LASSO回归构建相关预后模型。通过蛋白质-蛋白质相互作用(protein-protein interaction，PPI)网络筛选出核心的10个靶点。将模型基因和核心靶点进行交集。通过药物-化合物-靶点网络筛选出核心的化合物。将核心的化合物和靶点进行分子对接验证。构建列线图用以评估患者的预后情况。结果 得到片仔癀靶点162个，肝细胞癌靶点522个，片仔癀治疗肝细胞癌靶点20个，炎症相关靶点16个。GO和KEGG富集分析显示其发挥作用主要通过单加氧酶活力功能、氧化还原酶活性等生物学功能，化学致癌-受体激活等通路发挥作用。预后模型ROC曲线计算AUC分别为1年0.780，3年0.688，5年0.642，说明模型具有可靠性。预后模型与PPI核心靶点相交得到5个靶点：PON1、IGF2、NQO1、CCNB1和IGFBP3。列线图使用CCNB1、NQO1和T分期进行构建，其c-index为0.726，说明了模型的可靠性。药物-化合物-靶点网络提示槲皮素是核心的化合物，并且靶向上述2个基因。结论 片仔癀治疗肝细胞癌主要通过槲皮素靶向CCNB1和NQO1发挥抗炎作用，其预后模型可用于患者的生存预测。

Abstract:

OBJECTIVE To analyze the inflammation-related molecular targets of Pien Tze Huang in the treatment of hepatocellular carcinoma and to preliminary explore its mechanism. METHODS Obtain the ingredients and targets of Pien Tze Huang through TCMSP and BATMAN databases. Obtain the disease targets of hepatocellular carcinoma through Genecards, OMIM and TCGA databases. Take the intersection of compound targets and disease targets to get Pien Tze Huang’s target for the treatment of hepatocellular carcinoma. Obtain the related genes of inflammation pathway from the GSEA database, and then analyze the correlation between Pien Tze Huang’s therapeutic targets for hepatocellular carcinoma and inflammation-related genes to screen out inflammation-related targets, and explore the mechanism through GO and KEGG enrichment analysis. Then, single-factor cox analysis and LASSO regression were performed to construct related prognostic models. The 10 core targets were screened out through the protein-protein interaction(PPI) network. The model gene and the core target were intersected. The core compounds were screened out through the drug-compound-target network. Perform molecular docking verification between the core compound and the target. Construct a nomogram to assess the prognosis of patients. RESULTS Obtained 162 Pien Tze Huang targets, 522 hepatocellular carcinoma targets, 20 Pien Tze Huang therapeutic targets for hepatocellular carcinoma, and 16 inflammation-related targets. The enrichment analysis of GO and KEGG showed that their effects were mainly through biological functions such as monooxygenase activity, oxidoreductase activity, and chemical carcinogenesis-receptor activation. The ROC curve of the prognosis model calculated AUC as 0.780 in 1 year, 0.688 in 3 years, and 0.642 in 5 years, indicating that the model was reliable. The prognostic model intersects with the core target of PPI to get 5 targets: PON1, IGF2, NQO1, CCNB1 and IGFBP3. The nomogram was constructed using CCNB1, NQO1, and T staging, and its c-index was 0.726, indicating the reliability of the model. The drug-compound-target network suggested that quercetin was the core compound and targets the above two genes. CONCLUSION Pien Tze Huang’s treatment of hepatocellular carcinoma mainly uses quercetin to target CCNB1 and NQO1 to exert anti-inflammatory effects, and its prognostic model can be used to predict the survival of patients.