Screening and bioinformatics analysis of characteristics genes of sepsis and aging based on GEO database

Objective To screen the shared genes of sepsis and senescence by combining gene expression database (GEO) and machine learning algorithms, and to conduct bioinformatics analysis to understand the comorbid mechanism of sepsis and aging.Methods Sepsis-related genes were obtained from CTD, DisGeNet, and GeneCards, and aging-related genes were obtained from Aging Atlas database. GSE13904, GSE28705, and GSE8121 sepsis microarray data and aging dataset GSE173608 were obtained from the GEO database; Webstalt database was used for gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis. Gene Set Enrichment Analysis (GSEA) and Gene Set Variation Analysis (GSVA) were performed by CNSknowall, and immune cell infiltration analysis of the expression matrix of shared genes was performed using the CIBERSORTX platform. Finally, the BIDOS database was used to analyze the correlation between the expression level of characteristic genes and the survival time and APACHEII score of sepsis patients.Results There were 151 genes shared between sepsis and senescence, and the characteristic genes of TXN, CCL4, IL7 and SIRT1 were the common diagnostic markers of sepsis and senescence. The main gene sets that were up-regulated by the shared genes included cAMP signaling pathway, chemokine signaling pathway, PD-1 expression and PD-1 cancer checkpoint pathway, phospholipase D signaling pathway, PI3K-Akt signaling pathway, prolactin signaling pathway and other signaling pathways. Immune filtration analysis showed that the high expression of CCL4 was positively correlated with the prognosis of sepsis, and CCL4 was significantly positively correlated with activated natural killer cells. TXN was significantly positively correlated with resting dendritic cells.Conclusion CCL4, TXN, IL7 and SIRT1 can be used as diagnostic biomarkers for sepsis and aging, and are closely related to the pathophysiological process of sepsis and aging.

研究目的 本研究结合基因表达数据库（Gene Expression Omnibus，GEO）与机器学习算法筛选脓毒症与衰老的共有基因，并开展生物信息学分析以阐明二者的共病机制。研究方法 从比较毒物基因组学数据库（Comparative Toxicogenomics Database，CTD）、疾病基因关联数据库（DisGeNet）与基因卡片数据库（GeneCards）获取脓毒症相关基因，从衰老图谱数据库（Aging Atlas）获取衰老相关基因。从GEO数据库获取脓毒症芯片数据集GSE13904、GSE28705、GSE8121以及衰老数据集GSE173608；使用Webstalt数据库开展基因本体（Gene Ontology，GO）与京都基因与基因组百科全书（Kyoto Encyclopedia of Genes and Genomes，KEGG）富集分析。通过CNSknowall进行基因集富集分析（Gene Set Enrichment Analysis，GSEA）与基因集变异分析（Gene Set Variation Analysis，GSVA），使用CIBERSORTX平台对共有基因的表达矩阵开展免疫细胞浸润分析。最终借助BIDOS数据库分析特征基因表达水平与脓毒症患者生存时间及急性生理学与慢性健康状况评分II（APACHEII）的相关性。研究结果 脓毒症与衰老共存在151个共有基因，TXN、CCL4、IL7及SIRT1为二者的共同诊断标志物。共有基因上调富集的主要基因集包括cAMP信号通路、趋化因子信号通路、PD-1表达及PD-1癌症检查点通路、磷脂酶D信号通路、PI3K-Akt信号通路、催乳素信号通路等。免疫浸润分析显示，CCL4高表达与脓毒症预后呈正相关，且CCL4与活化自然杀伤细胞显著正相关；TXN与静息树突状细胞显著正相关。研究结论 TXN、CCL4、IL7及SIRT1可作为脓毒症与衰老的诊断生物标志物，且与二者的病理生理过程密切相关。