基于机器学习探讨不同年龄阶段半月板损伤铁死亡作用机制和免疫浸润研究

Exploring the mechanisms of ferroptosis and immune infiltration in meniscus injuries at different age stages using machine learning

目的:确定不同年龄阶段半月板损伤的重要铁死亡相关生物标志物和免疫浸润研究。方法:(1)从GEO数据库中获得训练数据集,使用GSE191157基因芯片表达矩阵分析数据,同时,从铁死亡官方下载相关基因进行进一步分析。(2)使用“limma”包对半月板损伤的老年患者和年轻患者进行差异表达基因分析,对GSE191157基因芯片的所有基因进行WGCNA分析。(3)选取2个与半月板损伤疾病最相关的颜色模块,再将相关模块的基因与铁死亡相关基因取交集。(4)随后将交集基因进行本体(GO)、京都基因和基因组百科全书(KEGG)分析。(5)通过机器学习LASSO回归对不同年龄半月板损伤铁死亡交集基因进行分析,提取核心基因。(6)最后进行基因集富集分析(GSEA)、SSGEVA免疫浸润等功能分析,确定核心基因与半月板损伤铁死亡确有相关性。结果:共鉴定出2517个差异基因,与铁死亡相关基因取交集得到61个基因,经WGCNA分析后得到与疾病相关性最高的青色和黄色模块,与之取交集得到46个交集基因;通过机器学习LASSO回归筛选出4个核心基因BEX1、MMP13、PTPN18和SLC38A1;GSEA分析结果提示半月板损伤铁死亡主要与Toll样受体信号通路、RNA转录与剪接、mTOR信号通路和ECM-受体相互作用有关;免疫细胞浸润分析显示出核心基因表达与免疫细胞浸润之间具有明显相关性,MMP13在效应记忆CD8 T细胞、CD56dim自然杀伤细胞、中央记忆CD4 T细胞中呈现负相关,而SLC38A1、BEX1、PTPN18三个核心基因呈正相关。结论:该研究共筛选出4个与不同年龄阶段半月板损伤的铁死亡相关关键基因,分别为BEX1、MMP13、PTPN18和SLC38A1;免疫细胞浸润分析显示出核心基因表达与免疫细胞浸润之间具有明显相关性。

Objective:To identify key ferroptosis-related biomarkers and investigate immune infiltration in meniscus injuries among individuals of different age groups.Methods:(1)We obtained our training dataset from the GEO database and analyzed gene expression matrices from the GSE191157 gene chip dataset.Simultaneously,we downloaded relevant genes associated with ferroptosis for further analysis.(2)Using the"limma"package,we conducted differential gene expression analysis between elderly and young patients with meniscus injuries and performed weighted gene co-expression network analysis(WGCNA)on all genes in the GSE191157 gene chip dataset.(3)We selected two color modules most correlated with meniscus injury and then intersectedthe genes in these modules with ferroptosis-related genes.(4)Subsequently,we conducted ontology(GO)and Kyoto Encyclopediaof Genes and Genomes(KEGG)analyses on the intersected genes(.5)Using LASSO regression in machine learning,we analyzedthe intersected genes related to ferroptosis in meniscus injuries among individuals of different ages and extracted core genes.(6)Finally,we conducted gene set enrichment analysis(GSEA)and single-sample gene set enrichment variation analysis(ssGSEA)for immune infiltration and confirmed the correlation between core genes and ferroptosis in meniscus injuries.Results:A totalof 2517 differentially expressed genes were identified,and the intersection of these genes with ferroptosis-related genes yielded61 genes.After WGCNA analysis,the blue and yellow modules were found to be most relevant to the disease,resulting in 46 intersectedgenes.Through LASSO regression analysis,four core genes(BEX1,MMP13,PTPN18,and SLC38A1)were selected.GSEA results indicated that ferroptosis in meniscus injuries is primarily associated with Toll-like receptor signaling,RNA transcriptionand splicing,mTOR signaling,and ECM-receptor interaction.Immune cell infiltration analysis revealed a significant correlationbetween core gene expression and immune cell infiltration.MMP13 exhibited a negative correlation with effector memoryCD8 T cells,CD56dim natural killer cells,and central memory CD4 T cells,while SLC38A1,BEX1,and PTPN18 were positivelycorrelated.Conclusion:This study identified four key genes associated with ferroptosis in meniscus injuries among individualsof different age groups,namely BEX1,MMP13,PTPN18,and SLC38A1.Immune cell infiltration analysis demonstrated aclear correlation between core gene expression and immune cell infiltration.