急性髓系白血病mRNA与非编码RNA差异表达谱及CeRNA调控网络分析

利用GEO数据库(gene expression omnibus database)通过生物信息学分析方法探讨急性髓系白血病(acute myelogenous leukemia，AML)的发病机制。检索GEO数据库中AML相关芯片数据集GSE142698、GSE142699和GSE96535。利用GEO2R分析得到差异mRNAs、miRNAs以及差异lncRNAs。利用在线生物信息学分析工具DAVID对差异mRNAs进行GO富集分析和KEGG通路分析。利用miRWalk数据库预测AML相关miRNAs的靶向mRNAs，利用Spongescan数据库预测AML相关miRNAs的靶向lncRNAs，构建lncRNA-miRNA-mRNA竞争性内源RNA （competing endogenous RNA,ceRNA）调控网络。共筛选出29个显著差异mRNAs、70个显著差异miRNAs和20 005个显著差异lncRNAs。GO富集分析和KEGG通路分析显示，差异表达基因主要涉及蛋白磷酸化、细胞分裂、细胞增殖的负调控、基因表达的正向调节、周期蛋白依赖的丝氨酸/苏氨酸激酶活性的调节等生物过程以及细胞周期、细胞衰老、癌症通路、PI3K-Akt通路等信号通路。将miRWalk数据库预测的靶向mRNAs与差异mRNAs取交集，Spongescan数据库预测的靶向lncRNAs与差异lncRNAs取交集，分别确定了25个mRNAs、6个lncRNAs参与AML相关ceRNA调控网络的构建。结果表明，lncRNAs可能作为关键的ceRNA，通过调控miRNA和相关靶基因参与AML的发生与发展，研究结果为AML诊断和治疗的分子生物学研究提供了新的依据。

Differential Expression Profiles of mRNA and Non-coding RNA and CeRNA Regulatory Network Analysis of Acute Myeloid Leukemia

The pathogenesis of acute myeloid leukemia （AML） was investigated by GEO database and bioinformatics analysis. AML related chip datasets GSE142698， GSE142699 and GSE96535 were retrieved from GEO database. Differential mRNAs， miRNAs and lncRNAs were obtained by GEO2R analysis. GO enrichment analysis and KEGG pathway analysis were performed for differential mRNAs using online bioinformatics analysis tool DAVID. The targeted mRNAs of AML related miRNAs were predicted by miRWalk database， and the targeted lncRNAs of AML related miRNAs were predicted by Spongescan database， and the lncRNa-miRNA-mRNA competing endogenous RNA （ceRNA） regulatory network was constructed. A total of 29 mRNAs， 70 miRNAs and 20 005 lncRNAs with significant differences were screened. GO enrichment and KEGG pathway analysis in differentially expressed genes mainly related to protein phosphorylation， cell division， negative regulation of cell proliferation， protein positive regulation of gene expression， cycle depends on serine/threonine kinase activity such as biological processes and the regulation of cell cycle， cell aging and cancer pathways， PI3K-Akt signaling pathways and other signaling pathways. The intersection of targeted mRNAs and differential mRNAs predicted by miRWalk database and targeted lncRNAs predicted by Spongescan database was selected. 25 mRNAs and 6 lncRNAs were identified to participate in the construction of AML related ceRNA regulatory networks. LncRNAs may be the key ceRNAS involved in the occurrence and development of AML by regulating miRNAs and related target genes， which may provide a new basis for molecular biological studies on the diagnosis and treatment of AML.