中华中医药学刊2025，Vol.43Issue(7)：80-85,后插29-后插35,13.DOI:10.13193/j.issn.1673-7717.2025.07.018
基于生物信息学与机器学习的坐骨神经痛与内质网应激相关生物标志物筛选及调控中药预测
Screening of Biomarkers Related to Sciatica and Endoplasmic Reticulum Stress Based on Bioinformatics and Machine Learning and Prediction of Regulating Chinese Medicine
摘要
Abstract
Objective This study aimed to screen the diagnostic biomarkers in the pathogenesis of sciatica and analyze the genes related to endoplasmic reticulum stress in the pathogenesis of sciatica.Methods The data of sciatica microarray was re-trieved from Geo database,and the differentially expressed genes(DEGs)associated with endoplasmic reticulum stress were ob-tained by analyzing the intersection of DEGs and endoplasmic reticulum stress genes.The differentially expressed genes were en-riched and analyzed,and the differentially expressed genes were analyzed by immune cell infiltration.The differentially expressed genes associated with immunity were further obtained by combining weighted gene coexpression network analysis.Key genes were screened by Lasso,xgboost and random forest methods to construct a diagnostic model and validate it.It objectively analyzed the correlation between endoplasmic reticulum stress related genes and diagnostic genes,and then predicted the Chinese medicine that can intervene the pathogenesis of sciatica and endoplasmic reticulum stress related genes.Results A total of 286 DEGs were ob-tained by differential analysis,which were mainly concentrated in biological processes such as cell response to reactive oxygen species,cell response to chemical stress,immune receptor activity,oxidoreductase activity,as well as peroxisomeproliferater-acti-vated receptor(PPAR)signaling pathway,phospholipase D signaling pathway,necrotic apoptosis and other signaling pathways.Immunoinfiltration analysis showed that regulatory T cells,follicular helper T cells and memory CD4 T cells were differentially ex-pressed,which also showed that the above three kinds of immune cells played a certain role in the pathogenesis of sciatica.A total of 83 key module genes were obtained by WGCNA combined with immune infiltration analysis.Hub genes such as ZC3H15,RPL7,RPL2611,TPT1,RPL2211,RPS15A,RPS3A,RPL36A,RPS24 and RPL30 were obtained by PPI analysis.Four diagnostic genes such as KLF9,scarf1,CYP27A1 and CEBPA were screened by machine learning method,and the four diagnostic genes were correlated with the expression of hub gene.Finally,36 kinds of Chinese medicine that can regulate hub gene were predicted.Conclusion KLF9,SCARF1,CYP27A1 and CEBPA genes have diagnostic value for sciatica.Endoplasmic reticulum stress related genes play a certain role in the pathogenesis of sciatica.Quanxie(Scorpio),Yanjingshe(cobra),Ziheche(dried human placen-ta),Liuhuang(sulfur),Lurong(Cervi Cornu Pantotrichum)and other Chinese medicines may be potential drugs for the preven-tion and treatment of sciatica.关键词
坐骨神经痛/中药/机器学习/生物信息学/内质网应激Key words
sciatica/Chinese medicine/machine learning/bioinformatics/endoplasmic reticulum stress分类
医药卫生引用本文复制引用
韦业,陈广辉,覃小伶,蒙映彤,邓豪,祁祥,何育风..基于生物信息学与机器学习的坐骨神经痛与内质网应激相关生物标志物筛选及调控中药预测[J].中华中医药学刊,2025,43(7):80-85,后插29-后插35,13.基金项目
国家自然科学基金项目(82160943) （82160943）
广西重点研发计划项目(桂科AB20159026) （桂科AB20159026）
广西自然科学基金项目(2020GXNSFAA259013) （2020GXNSFAA259013）
广西中医药重点培育学科建设项目(GZXK-Z-20-32) （GZXK-Z-20-32）
广西壮族自治区中医药管理局项目(GZZC2020042) （GZZC2020042）
广西中医药适宜技术开发与推广项目(GZSY21-13) （GZSY21-13）
广西中医药大学高层次人才培育创新团队项目(2022A006) （2022A006）
