【摘要】 目的:通过基因表达谱(GEO)数据库和癌症基因组图谱(TCGA)数据库筛选儿童低级别胶质瘤(LGG)的预后显著差异基因,为寻找儿童LGG预后生物标志物提供参考。方法:从GEO、TCGA数据库下载儿童LGG全基因组表达谱数据,筛选出儿童LGG差异表达基因。应用R语言进行基因GO、KEGG富集分析,绘制蛋白质间互作网络图,获取核心基因。采用单因素、多因素cox回归模型筛选与预后相关的显著差异基因。TISIDB数据库进行显著差异基因的生存分析,在线数据库GEPIA验证显著差异基因的表达。结果:GEO查找到一个数据集(GSE60898),TCGA获得174个样本数据。GEO数据库筛选出6 146个差异基因,TCGA数据库筛选出24 176个差异基因,GEO和TCGA两个数据库得到出共同表达的上调差异基因51个,下调差异基因20个。差异共表达的基因主要富集在粘着斑、ECM受体相互通路等信号通路。COX回归模型筛选出预后显著差异基因4个,分别是脂肪细胞增强子结合蛋白1(AEBP1)、纤溶酶原激活物抑制剂-1(SERPINA1)、胰岛素样生长因子结合蛋白2(IGFBP2)和维甲酸受体应答2(RARRES2)。预后分析结果表明显著差异基因与LGG的预后具有相关性(P<0.05),AEBP1、GFBP2、SERPINA1在肿瘤组织中高表达,在正常组织中低表达(P<0.05),而RARRES2在正常组织中低表达,在肿瘤组织中高表达(P<0.05)。结论:AEBP1、GFBP2、SERPINA1、RARRES2与LGG的预后具有相关性,可能是儿童LGG预后的潜在生物标志物。更多还原

【Abstract】 Objective: The gene expression profile(GEO) database and the Cancer Genome Atlas(TCGA) database were used to screen the prognostic significantly different genes of low-grade glioma(LGG) in children, and provide a reference for finding the prognostic biomarkers of LGG in children. Methods:The whole genome expression profile data of children’s LGG were downloadedfrom GEO and TCGA databases to screen out the differentially expressed genes of children’s LGG. Gene GO and KEGG enrichment analysis were performed by R language, and protein interaction network diagram was plottedto obtain the core genes. Single-factor and multi-factor cox regression models were used to screen the significantly different genes related to prognosis.The TISIDB database performed survival analysis of significantly different genes,and the online database GEPIA verified the expression of significantly different genes. Results: GEO found a data set(GSE60898), and TCGA obtained 174 sample data. The GEO database screened 6,146 differential genes, the TCGA database screened 24,176 differential genes, and the GEO and TCGA databases obtained 51 up-regulated differential genes and 20 down-regulated differential genes that were co-expressed.Differentially co-expressed genes were mainly enriched in signal pathways such as focal adhesion and ECM receptor interaction pathways. The COX regression model screened out 4 genes with significant prognostic differences, which were adipocyte enhancer binding protein 1(AEBP1), plasminogen activator inhibitor-1(SERPINA1), insulin-like growth factor binding protein 2(IGFBP2) and retinoic acid receptor response 2(RARRES2). The results of prognosis analysis showed that the differential genes were significantly correlated with the prognosis of LGG(P<0.05). AEBP1 I, GFBP2, SERPINA1 were highly expressed in tumor tissues, but low expressed in normal tissues(P<0.05), while RARRES2 was low expressedin normal tissues, and high expressed in tumor tissues(P<0.05). Conclusion: AEBP1 I, GFBP2, SERPINA1, RARRES2 are correlated with the prognosis of LGG, and may be potential biomarkers for the prognosis of LGG in children.更多还原