| 引用本文： |
-
莫泳锋,谢远亮,陈蔓莹,覃洪宇,叶雨.基于生物信息学构建前列腺癌单细胞层面lncRNA-miRNA-mRNA调控网络[J].广西科学,2022,29(3):462-475. [点击复制]
- MO Yongfeng,XIE Yuanliang,CHEN Manying,QIN Hongyu,YE Yu.Construction of a Single Cell Level lncRNA-miRNA-mRNA Regulatory Network of Prostatic Cancer Based on Bioinformatics[J].Guangxi Sciences,2022,29(3):462-475. [点击复制]
|
|
| 摘要: |
| 利用生物信息学筛选差异表达基因（Differentially Expressed Genes，DEGs），构建前列腺癌（Prostate Cancer，PCa）单细胞层面lncRNA-miRNA-mRNA调控网络，在分子水平上研究前列腺癌的发生发展过程及预后指标。首先从基因表达综合数据库（GEO）和癌症基因组图谱（TCGA）数据库分别下载单细胞RNA测序数据GSE157703和转录组测序数据TCGA-PRAD，通过R语言筛选差异表达信使RNA （DEmRNA）和差异表达长非编码RNA （DElncRNA），利用相关软件预测并构建lncRNA-miRNA-mRNA调控网络；然后使用R语言进行基因本体论（Gene Ontology，GO）和京都基因与基因百科全书（Kyoto Encyclopedia of Genes and Genomes，KEGG）分析，运用STRING数据库、Cytoscape软件建立蛋白相互作用网络，同时对单细胞RNA测序数据进行质控、降维、聚类和分群，构建单细胞层面的lncRNA-miRNA-mRNA调控网络；最后进行生存分析，通过人类蛋白质图谱进行验证。结果显示，共得到3 013个DEmRNA和1 101个DElncRNA，筛选出51个lncRNA、27个miRNA和97个mRNA构建lncRNA-miRNA-mRNA调控网络。GO分析揭示靶基因在腺状细胞迁移、化学突触传递的调节等生物学过程富集，KEGG分析揭示miRNA在癌症、轴突导向和胞间的黏附等信号通路富集。通过PPI得到Top 10基因，据此整合单细胞水平上PCa单细胞转录组测序数据，得到包括肿瘤细胞在内的9个细胞群的多条精确的lncRNA-miRNA-mRNA调控轴。生存分析结果显示，ITGA2、PDLIM5H和PRRG4低表达组的无病生存期显著高于高表达组（P<0.05），免疫组织化学验证了上述基因在肿瘤组织中均有不同程度的表达。本研究成功构建了PCa单细胞水平的lncRNA-miRNA-mRNA调控网络，并发现其不仅参与PCa的发生和发展，同时在临床预后预测方面也有重要意义。 |
| 关键词: 前列腺癌 lncRNA-miRNA-mRNA调控网络 单细胞测序 差异表达基因 生物信息学 |
| DOI：10.13656/j.cnki.gxkx.20220720.009 |
| 投稿时间：2022-03-30 |
| 基金项目:国家自然科学基金项目（82160501）资助。 |
|
| Construction of a Single Cell Level lncRNA-miRNA-mRNA Regulatory Network of Prostatic Cancer Based on Bioinformatics |
|
MO Yongfeng1, XIE Yuanliang2, CHEN Manying1, QIN Hongyu1, YE Yu1
|
| (1.The Second Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi, 530007, China;2.Tumor Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi, 530021, China) |
| Abstract: |
| Differentially Expressed Genes (DEGs) were screened by bioinformatics, and the lncRNA-miRNA-mRNA regulatory network at the single cell level of Prostate Cancer (PCa) was constructed to study the occurrence, development and prognosis of prostate cancer at the molecular level. Firstly, single cell RNA sequencing data GSE157703 and transcriptome sequencing data TCGA-PRAD were downloaded from Gene Expression Omnibus (GEO) and the Cancer Genome Atlas (TCGA) databases, respectively. Differentially Expressed messenger RNA (DEmRNA) and Differentially Expressed long non-coding RNA (DElncRNA) were screened by R language, and lncRNA-miRNA-mRNA regulatory network was predicted and constructed by using relevant software.Then Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed using R language, and the protein interaction network was established by STRING database and Cytoscape software. While the lncRNA-miRNA-mRNA regulatory network at the single cell level was constructed by quality control, dimensionality reduction, clustering and clustering of single cell RNA sequencing data.Finally, survival analysis was performed and validated by human protein map. The results showed that a total of 3 013 DEmRNA and 1 101 DElncRNA were obtained, and 51 lncRNA, 27 miRNA and 97 mRNA were screened to construct the lncRNA-miRNA-mRNA regulatory network. GO analysis revealed the enrichment of target genes in biological processes such as glandular cell migration and regulation of chemical synaptic transmission. KEGG analysis revealed the enrichment of miRNA in signaling pathways such as cancer, axon orientation and intercellular adhesion.Top 10 genes were obtained by PPI. Accordingly, PCa single cell transcriptome sequencing data were integrated to obtain multiple precise lncRNA-miRNA-mRNA regulatory axes at the single cell level for nine cell populations, including tumor cells. The results of survival analysis showed that the disease-free survival in the low-expression groups of ITGA2, PDLIM5H and PRRG4 was significantly higher than that in the high-expression groups (P<0.05). Immunohistochemistry confirmed that the above genes were expressed in different degrees in tumor tissues. In this study, the lncRNA-miRNA-mRNA regulatory network at the single cell level of PCa was successfully constructed, and found that it was not only involved in the occurrence and development of PCa, but also had important significance in clinical prognosis prediction. |
| Key words: prostate cancer lncRNA-miRNA-mRNA regulatory network single-cell sequencing differentially expressed genes bioinformatics |
