温脾通络开窍方调节铁死亡治疗阿尔茨海默病的分子机制

Study on the molecular mechanism of Wenpi Tongluo Kaiqiao Prescription on regulating ferroptosis in the treatment of Alzheimer’s disease

目的:利用生物信息学、网络药理学、免疫浸润分析和机器学习方法,探讨温脾通络开窍方调节铁死亡治疗阿尔茨海默病(AD)的可能分子机制。方法:通过网络药理学方法获取“温脾通络开窍方-AD-铁死亡”潜在的共同分子靶点。基于GSE33000数据集筛选共同分子靶点的差异表达基因(DEGs)并分析其相关性、京都基因与基因组百科全书(KEGG)通路富集情况和免疫浸润状态;对GSE33000的样本进行聚类分型;对亚型免疫细胞进行差异分析;构建机器学习模型筛选DEGs关键基因;构建风险预测列线图模型并验证;对关键基因进行临床相关性分析。结果:共获得15个“温脾通络开窍方-AD-铁死亡”共同分子靶点,其中13个是DEGs。13个DEGs具有相互调控、介导MAPK等信号通路和免疫调节作用,且和免疫细胞在各亚型AD患者分布有所差异。GLM机器学习模型的预测性能最高,其重要性评分前5名的关键基因为TP53、HSPB1、RELA、MYC和NFE2L2。其中,HSPB1与AD患者年龄负相关。列线图预测模型准确性较高,预测误差风险较小。结论:温脾通络开窍方调节铁死亡治疗AD可能与TP53、HSPB1、RELA等分子靶点及其介导的MAPK等信号通路和免疫调节密切相关。基于分子靶点的GLM模型可较准确诊断AD,亦可间接评估温脾通络开窍方治疗AD的效能。

Objective:To investigate the possible molecular mechanism of Wenpi Tongluo Kaiqiao Prescription regulating ferroptosis in the treatment of Alzheimer’s disease(AD)by using bioinformatics,network pharmacology,immunoinfiltration analysis and machine learning methods.Methods:The potential common molecular targets of‘Wenpi Tongluo Kaiqiao Prescription-AD-ferroptosis’were obtained by network pharmacological method.Based on the GSE33000 data set,differentially expressed genes(DEGs)at common molecular targets were selected,and.the correlation,enrichment of Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway,and immunoinfiltration status were analyzed.The samples of GSE33000 were clustered and classified.The difference of subtype immune cells was analyzed.Construct machine learning model to screen key genes of DEGs;Construct and verify the risk prediction line graph model;Clinical correlation analysis of key genes was performed.Results:A total of 15 common molecular targets of‘Wenpi Tongluo Kaiqiao Prescription-AD-ferroptosis’were obtained,13 of which were DEGs.The common molecular target DEGs has the function of mutual regulation,mediating MAPK and other signaling pathways and immunomodulatory,and the distribution of DEGs and immune cells is different in all subtypes of AD patients.The prediction performance of GLM machine learning model was the highest,and the top 5 key genes in its importance score were TP53,HSPB1,RELA,MYC and NFE2L2.Among them,HSPB1 was negatively correlated with the age of AD patients.The prediction model of line graph has high accuracy and low risk of prediction error.Conclusion:The regulation of AD by Wenpi Tongluo Kaiqiao Prescription in the treatment of ferroptosis may be closely related to molecular targets such as TP53,HSPB1 and RELA,and their mediated signaling pathways such as MAPK and immune regulation.The molecular targetbased GLM model can accurately diagnose AD and indirectly evaluate the efficacy of Wenpi Tongluo Kaiqiao Prescription in the treatment of AD.