基于机器学习鉴别风寒湿痹型类风湿性关节炎自噬相关生物标志物和免疫浸润及中药预测

Machine learning-based identification of autophagy-related biomarkers, immune infiltration in wind-cold-dampness pattern rheumatoid arthritis, and traditional Chinese medicine prediction

背景:类风湿性关节炎是一种慢性自身免疫性疾病,以对称性关节炎症、关节损害和全身症状为特征。目前自噬在风寒湿痹型类风湿关节炎中的作用和机制仍不清楚。本研究旨在识别风寒湿痹型类风湿性关节炎中潜在的自噬相关标志物,并探讨类风湿性关节炎滑膜组织中自噬相关基因和免疫微环境的作用和机制。目的:鉴别风寒湿痹型类风湿性关节炎自噬相关生物标志物和免疫浸润情况。方法:从GEO数据库获得正常和风寒湿痹型类风湿性关节炎基因表达谱,通过差异分析得到差异基因,从人类自噬数据库获得自噬相关基因,取交集,得到自噬差异基因。采用基因本体论(GO)、京都基因与基因组百科全书(KEGG)富集分析初步揭示其机制。使用机器学习和蛋白质-蛋白质相互作用网络识别类风湿性关节炎具有高度相关特征的自噬差异基因,通过基因集富集分析(Gene Set Enrichment Analysis,GSEA)进一步揭示风寒湿痹型类风湿性关节炎的作用机制。采用单样本GSEA算法研究类风湿性关节炎的免疫浸润特征及其与自噬相关基因的关系。通过在线网站进行中药预测。结果:正常和类风湿性关节炎滑膜样本中发现了24个自噬差异表达基因。根据功能富集分析,差异表达基因主要参与线粒体解聚、线粒体自噬、大范围自噬的正调控和自噬的调控。使用单样本GSEA算法,我们发现类风湿性关节炎样本中多种免疫细胞浸润和活化显著增加,并且与核心自噬差异基因密切相关。预测出65味中药可通过自噬相关基因靶点治疗疾病。结论:自噬可能通过诱导免疫炎症来促进类风湿性关节炎的进展,其中多种免疫细胞参与风寒湿痹型类风湿性关节炎进展,带电多泡体蛋白4B(Charged Multivesicular Body Protein 4B,CHMP4B)、线粒体转录终止因子结构域1(Mitochondrial Transcription Termination Factor Domain 1,MTERFD1)和突触体相关膜蛋白(Synaptosomal Associated Protein,SNAP)29可作为类风湿性关节炎生物标志物和潜在治疗靶点。

Background: Rheumatoid arthritis(RA) is a chronic autoimmune disease characterized by symmetrical joint inflammation, joint damage, and systemic symptoms. Currently, the role and mechanisms of autophagy in the wind-cold-dampness pattern of RA remain unclear. This study aims to identify potential autophagy-related biomarkers in wind-cold-dampness RA and explore the role and mechanisms of autophagy-related genes and the immune microenvironment in RA synovial tissue. Objective: To identify autophagy-related biomarkers and immune infiltration in RA. Methods: Normal and wind-cold-dampness RA gene expression profiles were obtained from the GEO database, and differential genes were obtained through differential analysis. Autophagy-related genes were retrieved from the human autophagy database, and the intersection was taken to obtain autophagy differential genes. Gene Ontology(GO) and Kyoto Encyclopedia of Genes and Genomes(KEGG) were used to preliminarily reveal their mechanisms. Machine learning and PPI network were employed to identify autophagy differential genes highly associated with RA features. GSEA analysis was performed to further elucidate the mechanisms of wind-cold-dampness RA. Single-sample GSEA algorithm was used to study the immune infiltration characteristics of RA and its relationship with autophagy-related genes. Traditional Chinese medicine prediction was conducted online. Results: Twenty-four DEGs were identified in normal and RA synovial samples. According to functional enrichment analysis, DEGs were mainly involved in the upregulation of mitochondrial disintegration, mitochondrial autophagy, broad-range autophagy, and the regulation of autophagy. Using the single-sample GSEA algorithm, we found a significant increase in various immune cell infiltrations and activations in RA samples, which was closely associated with core autophagy differential genes. Sixty-five traditional Chinese medicines that can treat diseases through autophagy-related gene targets were predicted. Conclusion: Autophagy may promote the progression of RA by inducing immune inflammation, with various immune cells participating in the progression of wind-cold-dampness RA. CHMP4B, MTERFD1, and SNAP29 can serve as RA biomarkers and potential therapeutic targets.