基于基因表达综合数据库的缺血性卒中缺氧相关差异基因表达分析

Expression analysis of hypoxic-related differentially expressed genes in ischemic stroke based on gene expression omnibus database

目的基于基因表达综合(GEO)数据库,采用生物信息学方法分析缺血性卒中的基因表达情况,结合缺氧相关基因,解析缺氧相关差异基因(HRDEGs)表达特征,筛选出关键基因,为深入了解缺血性卒中提供重要支撑。方法从GEO数据库下载GSE16561和GSE58294两个数据集,采用Python软件进行数据合并,利用Combat方法消除批次效应,同时保留疾病分组特征。对消除批次效应前后的数据采用主成分分析法进行降维处理,并对缺血性卒中组和正常对照组人群进行组内相关系数(ICC)检测。对合并及批次效应消除后数据集进行基因集富集分析(GSEA)及单样本GSEA,以名义P值(NOM P-val)<0.05且假阳性率P值(FDR P-val)<0.25作为标准,筛选出具有显著差异的基因集。利用R软件对GSE16561和GSE58294数据集合并及批次效应消除后的缺血性卒中患者和正常对照者之间的差异表达基因进行鉴定[以log 2基因表达差异倍数(FC)的绝对值≥0.58和矫正P值(P adj)<0.05作为筛选标准],并与在美国国立生物技术信息中心(NCBI)中获取的缺氧相关基因进行交集,得到HRDEGs。对HRDEGs进行基因本体论(GO)和京都基因与基因组百科全书(KEGG)富集分析,使用STRING数据库构建差异表达基因的蛋白质互相作用网络,利用Cytoscape3.8软件筛选排名前10位的关键基因。结果ICC分析结果显示,消除批次效应后的GSE16561和GSE58294数据集中缺血性卒中和正常对照者的ICC分别为0.94和0.98,一致性极好。GSEA结果显示,对GSE16561和GSE58294合并及批次效应消除后的新数据集中,34个基因集在缺血性卒中样本中显著富集,筛选出表达差异基因404个(均P adj<0.05),其中高表达基因354个,低表达基因50个。与缺氧相关基因进行交集,获得64个HRDEGs。GO富集分析表明,HRDEGs主要在囊泡腔、细胞质囊泡腔、分泌颗粒腔和特殊颗粒中显著富集,具有酰胺结合、肽结合、磷脂结合和酶抑制活性等分子功能,主要参与了细胞因子产生的正向调节、炎性反应的调节、对细菌来源分子的反应和对脂多糖的反应等生物学过程。KEGG富集分析表明,HRDEGs主要在血脂与动脉粥样硬化、沙门氏菌感染、中性粒细胞胞外陷阱形成、核苷酸结合寡聚化结构域样受体信号通路、癌症中的蛋白聚糖、结核病和坏死性凋亡等通路上存在富集。基于蛋白质互相作用网络,最终筛选出了10个关键基因,分别为ARG1(精氨酸酶1)、CASP1(含半胱氨酸的天冬氨酸蛋白水解酶1)、IL-1R1(白细胞介素1受体1型)、ITGAM(整合素亚基αM)、MMP9(基质金属蛋白酶9)、PTGS2(前列腺素内过氧化物合酶2)、STAT3(信号传感器和转录激活剂3)、TLR2(Toll样受体2)、TLR4、TLR8。。结论通过生物信息学挖掘,本研究获得了10个缺血性卒中与缺氧相关的关键基因,可能为后续研究工作及诊断治疗提供了潜在靶点。

Objective Based on the gene expression omnibus(GEO)database,bioinformatics methods were employed to analyze the expression characteristics of hypoxia-related differentially expressed genes(HRDEGs)in ischemic stroke,and key genes were screened,to provide important support for a deeper understanding of ischemic stroke.Methods The GSE16561 and GSE58294 datasets were downloaded from the GEO database,and Python software was used for data integration.The Combat method was employed to eliminate batch effects while retaining disease grouping characteristics.Principal component analysis was conducted to reduce dimensionality of the data before and after batch effect removal,and intraclass correlation coefficient(ICC)testing was performed on the ischemic stroke and normal control groups.Gene set enrichment analysis(GSEA)and single-sample GSEA were conducted on the merged and batch effects eliminated dataset,with a nominal P-value(NOM P-val)<0.05 and false discovery rate P-value(FDR P-val)<0.25 used as criteria to select significantly different gene sets.Differential expression genes between the ischemic stroke samples and normal control samples after merging and eliminating batch effects of the GSE16561 and GSE58294 datasets were identified using R software,with an absolute value of log 2 gene expression fold change(FC)≥0.58 and adjusted P-value(P adj)<0.05 as selection criteria.Intersection with hypoxia-related genes obtained from the National Center for Biotechnology Information(NCBI)in the United States yielded the HRDEGs.Gene ontology(GO)and Kyoto encyclopedia of genes and genomes(KEGG)enrichment analyses were performed on the HRDEGs,and the STRING database was used to construct a protein-protein interaction network of differentially expressed genes.The top 10 key genes were filtered using Cytoscape 3.8 software.Results The ICC analysis results showed excellent consistency in the ischemic stroke and normal control samples after batch effect removal,with ICC values of 0.94 and 0.98 for the GSE16561 and GSE58294 datasets,respectively.GSEA results demonstrated significant enrichment of 34 gene sets in the stroke samples in the newly merged and batch effects removed dataset from GSE16561 and GSE58294,leading to the identification of 404 differentially expressed genes(all with P adj<0.05),including 354 upregulated genes and 50 downregulated genes.Intersection with hypoxia-related genes yielded 64 HRDEGs.GO enrichment analysis indicated significant enrichment of HRDEGs in vesicle lumen,cytoplasmic vesicle lumen,secretory granule lumen,with molecular functions such as amide binding,peptide binding,phospholipid binding,and enzyme inhibitor activity.These genes are primarily involved in the positive regulation of cytokine production,regulation of immune response,response to bacterium-derived molecules,and response to lipopolysaccharide,among other biological processes.KEGG enrichment analysis revealed enrichment of HRDEGs in pathways related to lipid and atherosclerosis,Salmonella infection,neutrophil extracellular trap formation,nucleotide-binding oligomerization domain-like receptor signaling pathway,protein glycosylation in cancer,tuberculosis,and necroptosis.Based on the protein-protein interaction network,10 key genes were identified,including arginase1(ARG1),caspase1(CASP1),interleukin1 receptor type 1(IL-1R1),integrin subunit alpha M(ITGAM),matrix metalloproteinase9(MMP9),prostaglandin-endoperoxide synthase 2(PTGS2),signal transducer and activator of transcription 3(STAT3),Toll-like receptor 2(TLR2),TLR4,and TLR8.Conclusion This study has identified 10 key genes associated with ischemic stroke and hypoxia through bioinformatics mining,which maybe provid potential targets for subsequent research and diagnostic and therapeutic interventions.