现代检验医学杂志2026，Vol.41Issue(2)：160-166,7.DOI:10.3969/j.issn.1671-7414.2026.02.027
基于网络毒理学和生物信息学探索双酚A对骨性关节炎发生的潜在影响及关键基因筛选的分析研究
Analytical Study of Bisphenol A Potential Impact on Osteoarthritis Development and Key Gene Screening Based on Network Toxicology and Bioinformatics
摘要
Abstract
Objective Network toxicology and bioinformatics methods were applied to investigate in-depth the potential ef-fects of bisphenol A(BPA)on the development of osteoarthritis(OA)and identify key genes in the process.Methods Three OA datasets were downloaded from the Gene Expression Omnibus(GEO)database at the National Center for Biotechnol-ogy Information(NCBI),and differentially expressed genes(DEGs)were identified using R software.The target genes for BPA-induced OA were obtained from the Comparative Toxicogenomics Database(CTD).The intersection of relevant genes from the two databases was identified using Venn diagrams,followed by Gene Ontology(GO)analysis,and Kyoto Encyclo-pedia of Genes and Genomes(KEGG)enrichment analysis.Six machine learning models(Lasso regression,support vector machine,Boruta algorithm,XGBoost,LightGBM,and AdaBoost methods)were employed to identify biomarkers for OA.Disease risk nomogram was further constructed,and its validity was assessed by calibration curves,clinical impact curves and ROC curves.The expression levels of the targets were validated.In addition,the binding potential between BPA and targets was assessed by molecular docking.Potential target drugs were identified from the DGIdb database.Expression of hub genes in OA and normal groups was verified by external datasets.Results After integrating the dataset,a total of 15 intersecting genes were identified.Enrichment analysis showed that these genes were associated with functional pathways such as mitotic nuclear division,receptor binding regulation,p53 signaling axis,IL-17 mediated activation pathways and TNF transduction mechanisms.Machine learning algorithms further identified IGFBP1,CDH2 and MKI67 as key genes.Receiver operating characteristic(ROC)curves demonstrated that these genes had high diagnostic efficacy(AUC=0.926)and were significantly overexpressed in the OA group.Expression levels of these three genes were externally validated in the GSE51588 dataset.Mo-lecular docking confirmed stable and robust binding between BPA and its targets.Finally,13 drugs targeting IGFBP1 were identi-fied along with methadone hydrochloride and methadone hydrochloride targeting CDH2,and identified abemaciclib and sunitinib targeting MKI67.These compounds represent potential therapeutic agents for OA treatment.Conclusions This study successful-ly elucidates the potential mechanism by which BPA influences OA development and identifies IGFBP1,CDH2 and MKI67 as three major biomarkers.These markers may serve as potential biomarkers and therapeutic targets for OA.关键词
网络毒理学/生物信息学/双酚A/骨性关节炎/机器学习Key words
network toxicology/bioinformatics/bisphenol A/osteoarthritis/machine learning分类
医药卫生引用本文复制引用
饶亚妮,黄新周,卫永鲲..基于网络毒理学和生物信息学探索双酚A对骨性关节炎发生的潜在影响及关键基因筛选的分析研究[J].现代检验医学杂志,2026,41(2):160-166,7.基金项目
国家临床重点专科建设项目[陕卫医函(2023)325号]. （2023）
