Abstract:

To screen potential key molecular biomarkers of corona virus disease 2019 （COVID-19） infection and analyze their immune infiltration characteristics by using bioinformatics methods. The GSE152418 dataset was obtained from the GEO database， including 17 COVID-19 patients and 17 healthy controls. The module genes most related to COVID-19 were screened out by weighted gene co-expression network analysis （WGCNA）， then intersected with differentially expressed genes （DEGs） to obtain common genes （CGs）. Perform function and signal pathway enrichment analysis of CGs， construct protein interaction （PPI） network to screen key （Hub） genes， construct miRNA-transcription factor （TF）-mRNA regulatory network of Hub genes， and use CIBERSORT algorithm to predict immune cells in samples infiltration characteristics. Difference analysis yielded 2 049 DEGs. Among the seven modules analyzed by WGCNA， the "turquoise" module had the highest correlation with COVID-19 （r=0.91， P<0.001）. The gene significance in the module and the module membership were significantly positively correlated （r=0.96， P<0.001）. A total of 766 CGs were obtained， which were mainly involved in mitosis， microtubule binding， cation channel activity， oocyte meiosis， and cell senescence. The Top 10 Hub genes screened by PPI network were CDK1， BUB1， CCNA2， CDC20， KIF11， BUB1B， CDCA8， TOP2A， CCNB2， KIF20A. The constructed miRNA-TF-mRNA network included 51 miRNAs， 5 TFs， 10 mRNA. Compared with the healthy control group， the infiltration levels of naive B cells and eosinophils in COVID-19 patients were decreased （P<0.05）， and the infiltration levels of plasma cells and activated mast cells were increased （P<0.05）. In this study， 10 Hub genes were screened through WGCNA and PPI network analysis， and 5 TFs and 51 miRNAs were predicted to regulate Hub genes. There were significant differences in immune infiltration characteristics between COVID-19 patients and healthy controls. These immune cells relevant molecular markers may serve as potential targets for COVID-19 immunotherapy.

Key words: COVID-19, biomarkers, immune infiltration, bioinformatics, infection

摘要：

应用生物信息学方法筛选新型冠状病毒肺炎（corona virus disease 2019，COVID-19）感染的潜在关键分子生物标志物并分析其免疫浸润特征。从GEO数据库下载GSE152418数据集，其中COVID-19患者17例，健康对照17例。用加权基因共表达网络分析（weighted gene co-expression network analysis，WGCNA）方法筛选出COVID-19最相关的模块基因。与差异基因取交集得到共同基因，进行功能及信号通路富集分析，构建蛋白互作网络筛选关键基因，构建关键基因的miRNA-TF-mRNA调控网络，用CIBERSORT算法预测样本免疫细胞浸润特征。差异分析得到2 049个差异基因。WGCNA分析7个模块中“土耳其蓝色”模块与COVID-19相关性最高（r=0.91，P<0.001）。模块中基因显著性和模块隶属度呈显著正相关（r=0.96，P<0.001）。得到共同基因766个，主要参与有丝分裂、微管结合、阳离子通道活性及卵母细胞减数分裂、细胞衰老等。蛋白互作网络筛选到前10位关键基因分别为CDK1、BUB1、CCNA2、CDC20、KIF11、BUB1B、CDCA8、TOP2A、CCNB2、KIF20A，构建的miRNA-TF-mRNA网络包含51个miRNA、5个TF、10个mRNA。COVID-19患者较健康对照组幼稚B细胞、嗜酸性粒细胞浸润水平显著降低（P<0.05），浆细胞、活化肥大细胞浸润水平显著升高（P<0.05）。通过WGCNA及蛋白互作网络分析筛选出10个关键基因，并预测到调控关键基因的5个TF及51个miRNA，且COVID-19患者与健康对照的免疫浸润特征存在统计学差异，这些与免疫细胞相关的分子标志物可能作为COVID-19免疫治疗的潜在靶标。

关键词: 新型冠状病毒肺炎, 生物标志物, 免疫浸润, 生物信息学, 感染