In silico analysis excavates potential biomarkers by constructing miRNA-mRNA networks between non-cirrhotic HCC and cirrhotic HCC
In silico analysis excavates potential biomarkers by constructing miRNA-mRNA networks between non-cirrhotic HCC and cirrhotic HCC
复制标题
计算机分析通过构建非肝硬化 HCC 和肝硬化 HCC 之间的 miRNA-mRNA 网络来挖掘潜在的生物标志物
DOI：
10.1186/s12935-019-0901-3
复制
发表时间：
2019-07-18
影响因子：
5.8
通讯作者：
Fan, Weimin
中科院分区：
文献类型：
--
作者：
Lt, Bisha Ding;Lou, Weiyang;Fan, Weimin
背景：越来越多的证据表明，伴或不伴肝硬化的 HCC 患者具有不同的临床特征、肿瘤发展和预后。然而，很少有研究直接研究非肝硬化HCC和肝硬化HCC之间的潜在分子机制。方法：从癌症基因组图谱（TCGA）数据库下载临床信息和RNA-seq数据。通过R软件获得有或没有肝硬化的HCC的差异表达基因（DEG）。 Enrichr 进行功能注释和通路富集分析。通过STRING建立蛋白质-蛋白质相互作用（PPI）网络，并将其映射到Cytoscape以识别中心基因。通过miRDB数据库预测MicroRNA。此外，通过starBase数据库对所选基因和miRNA之间进行相关性分析。通过 GEO 数据集进一步验证了伴或不伴肝硬化的 HCC 与相应正常肝组织之间的 miRNA 表达水平。最后，通过qRT-PCR验证关键miRNA和靶基因的表达水平。结果：TCGA中的132例非肝硬化HCC和79例肝硬化HCC之间，获得了768个DEG，主要涉及神经活性配体-受体相互作用途径。根据TCGA中基因表达分析结果，CCL19、CCL25、CNR1、PF4和PPBP被重命名为关键基因并选择进行进一步研究。生存分析表明，CNR1 上调与肝硬化 HCC 的 OS 较差相关。此外，ROC分析揭示了PF4和PPBP对肝硬化HCC以及CCL19、CCL25对非肝硬化HCC的显着诊断价值。接下来，预计有 517 个 miRNA 靶向 5 个关键基因。相关性分析证实，517 个 miRNA 中有 16 个对关键基因产生负调控。通过检测GEO数据库中这些关键miRNA的表达水平，我们发现4个miRNA具有较高的研究价值。最后，根据qRT-PCR的结果构建了潜在的miRNA-mRNA网络。结论：在计算机分析中，我们首先构建了非肝硬化HCC和肝硬化HCC中的miRNA-mRNA调控网络。
Background: Mounting evidences have demonstrated that HCC patients with or without cirrhosis possess different clinical characteristics, tumor development and prognosis. However, few studies directly investigated the underlying molecular mechanisms between non-cirrhotic HCC and cirrhotic HCC.Methods: The clinical information and RNA-seq data were downloaded from The Cancer Genome Atlas (TCGA) database. Differentially expressed genes (DEGs) of HCC with or without cirrhosis were obtained by R software. Functional annotation and pathway enrichment analysis were performed by Enrichr. Protein-protein interaction (PPI) network was established through STRING and mapped to Cytoscape to identify hub genes. MicroRNAs were predicted through miRDB database. Furthermore, correlation analysis between selected genes and miRNAs were conducted via starBase database. MiRNAs expression levels between HCC with or without cirrhosis and corresponding normal liver tissues were further validated through GEO datasets. Finally, expression levels of key miRNAs and target genes were validated through qRT-PCR.Results: Between 132 non-cirrhotic HCC and 79 cirrhotic HCC in TCGA, 768 DEGs were acquired, mainly involved in neuroactive ligand-receptor interaction pathway. According to the result from gene expression analysis in TCGA, CCL19, CCL25, CNR1, PF4 and PPBP were renamed as key genes and selected for further investigation. Survival analysis indicated that upregulated CNR1 correlated with worse OS in cirrhotic HCC. Furthermore, ROC analysis revealed the significant diagnostic values of PF4 and PPBP in cirrhotic HCC, and CCL19, CCL25 in non-cirrhotic HCC. Next, 517 miRNAs were predicted to target the 5 key genes. Correlation analysis confirmed that 16 of 517 miRNAs were negatively regulated the key genes. By detecting the expression levels of these key miRNAs from GEO database, we found 4 miRNAs have high research values. Finally, potential miRNA-mRNA networks were constructed based on the results of qRT-PCR.Conclusion: In silico analysis, we first constructed the miRNA-mRNA regulatory networks in non-cirrhotic HCC and cirrhotic HCC.