He Jin-Hua, Han Ze-Ping, Zou Mao-Xian, Wang Li, Lv Yu Bing, Zhou Jia Bin, Cao Ming-Rong, Li Yu-Guang
1 Department of Laboratory, Central Hospital of Panyu District , Guangzhou, China .
2 Department of General Surgery, First Affiliated Hospital, Jinan University , Guangzhou, China .
J Comput Biol. 2018 Feb;25(2):146-157. doi: 10.1089/cmb.2016.0093. Epub 2017 Aug 24.
Information processing tools and bioinformatics software have significantly advanced researchers' ability to process and analyze biological data. Molecular data from human and model organism genomes help researchers identify topics for study, which, in turn, improves predictive accuracy, facilitates the identification of relevant genes, and simplifies the validation of laboratory data. The objective of this study was to explore the regulatory network constituted by long noncoding RNA (lncRNA), miRNA, and mRNA in prostate cancer (PCa). Microarray data of PCa were downloaded from The Cancer Genome Atlas database and DESeq package in R language were used to identify the differentially expressed genes (DEGs) between PCa and normal samples. Gene ontology enrichment analysis of DEGs was conducted using the Database for Annotation, Visualization, and Integrated Discovery. TargetScan, microcosm, miRanda, miRDB, and PicTar were used to predict target genes. LncRNA associated with PCa was exploited in the lncRNASNP database, and the LncRNA-miRNA-mRNA regulatory network was visualized using Cytoscape. Our study identified 57 differentially expressed miRNAs and 1252 differentially expressed mRNAs; of these, 691 were downregulated genes primarily involved in focal adhesion, vascular smooth muscle contraction, calcium signaling pathway, and so on. The remaining 561 were upregulated genes principally involved in systemic lupus erythematosus, progesterone-mediated oocyte maturation, oocyte meiosis, and so on. Through the integrated analysis of correlation and target gene prediction, our studies identified 1214 miRNA:mRNA pairs, including 52 miRNAs and 395 mRNAs, and screened out 455 lncRNA-miRNA pairs containing 52 miRNAs. Therefore, owing to the interrelationship of lncRNAs and miRNAs with mRNAs, our study screened out 19,075 regulatory relationships. Our data provide a comprehensive bioinformatics analysis of genes, functions, and pathways that may be involved in the pathogenesis of PCa.
信息处理工具和生物信息学软件显著提升了研究人员处理和分析生物数据的能力。来自人类和模式生物基因组的分子数据有助于研究人员确定研究主题，进而提高预测准确性、促进相关基因的识别并简化实验室数据的验证。本研究的目的是探索长链非编码RNA（lncRNA）、微小RNA（miRNA）和信使RNA（mRNA）在前列腺癌（PCa）中构成的调控网络。从癌症基因组图谱数据库下载PCa的微阵列数据，并使用R语言中的DESeq软件包来识别PCa与正常样本之间的差异表达基因（DEG）。使用注释、可视化和综合发现数据库对DEG进行基因本体富集分析。利用TargetScan、microcosm、miRanda、miRDB和PicTar来预测靶基因。在lncRNASNP数据库中利用与PCa相关的lncRNA，并使用Cytoscape可视化lncRNA-miRNA-mRNA调控网络。我们的研究鉴定出57个差异表达的miRNA和1252个差异表达的mRNA；其中，691个是下调基因，主要参与粘着斑、血管平滑肌收缩、钙信号通路等。其余561个是上调基因，主要参与系统性红斑狼疮、孕酮介导的卵母细胞成熟、卵母细胞减数分裂等。通过相关性和靶基因预测的综合分析，我们的研究鉴定出1214个miRNA:mRNA对，包括52个miRNA和395个mRNA，并筛选出455个包含52个miRNA的lncRNA-miRNA对。因此，由于lncRNA和miRNA与mRNA的相互关系，我们的研究筛选出19,075个调控关系。我们的数据为可能参与PCa发病机制的基因、功能和通路提供了全面的生物信息学分析。
