目的：利用生物信息学方法分析与葡萄膜恶性黑色素瘤转移相关的非编码RNA，以及它们作为竞争性内源RNA的作用机制。方法：从癌症基因组图谱(The Cancer Genome Atlas，TCGA)数据库下载80例葡萄膜恶性黑色素瘤患者的RNA测序数据和临床资料，采用edgeR算法分析转移与非转移患者组织中差异表达(differentially expressed，DE)的长链非编码RNA(lncRNA)、微小RNA(miR)和mRNA，并构建lncRNA-miR-mRNA的竞争性内源RNA(competing endogenous RNA，ceRNA)调控网络，基因富集分析和通路分析研究网络中mRNA的生物学功能。Kaplan-Meier生存曲线分析ceRNA网络中核心RNA与生存率的关系。结果：从发生远处转移的葡萄膜恶性黑色素瘤样本中，共鉴定出346个上调的mRNA，118个下调的miR和45个上调的lncRNA。其中67个mRNA，7个miR和30个lncRNA相互组合形成616个ceRNA单元，并形成了一个具有181条边线ceRNA网络。基因富集分析表明：网络中的mRNA富集在肿瘤生成和转移相关的几个基因本体(Gene Ontology)和信号通路。拓扑分析确定了6个核心lncRNA(LINC00861、LINC02421、BHLHE40-AS1、LINC01252、LINC00513和LINC02389)和3个核心mRNA(UNC5D、BCL11B和MTDH)。 所有核心lncRNA、核心mRNA的表达水平和5个miR(miR-221、miR-222、miR-506、miR-507、miR-876)的表达水平均与总体生存率显着相关(均P<0.05)。结论：本研究揭示了几种lncRNA及其相关的ceRNA网络在葡萄膜恶性黑色素瘤转移中的作用，为进一步研究葡萄膜恶性黑色素瘤的发生和/或转移提供了新的方向。

Objective: To elucidate the expression of long non-coding RNAs (lncRNAs) and their roles as competing endogenous RNAs (ceRNAs) in uveal melanoma (UM) metastasis. Methods: RNA sequencing data and clinical information of 80 patients with UM were obtained from The Cancer Genome Atlas (TCGA) database. Differentially expressed (DE) mRNAs, microRNAs (miR), and lncRNAs between metastatic and non-metastatic individuals with UM were screened using the edgeR algorithm. Gene enrichment analysis was conducted for the DE mRNAs. LncRNA-miR-mRNA regulatory triples and a ceRNA network were constructed. Betweenness centrality was used to screen hub genes and lncRNAs for subnetwork analysis. Kaplan-Meier survival analysis was conducted to explore correlations between the expression of hub RNAs and overall survival in the TCGA UM cohort. Results: A total of 346 upregulated mRNAs, 118 downregulated miRs, and 45 upregulated lncRNAs were identified in samples with systemic metastasis. Among them, 67 mRNAs, 7 miRs, and 30 lncRNAs mapped to 616 ceRNA triples, thus forming an interconnected ceRNA network with 181 edges. Gene enrichment analysis revealed that mRNAs in the network were enriched in multiple gene ontology terms and pathways associated with carcinogenesis and metastasis. Topological analysis identified 6 hub lncRNAs (LINC00861, LINC02421, BHLHE40-AS1, LINC01252, LINC00513, and LINC02389) and 3 hub mRNAs (UNC5D, BCL11B, and MTDH). The expression levels of all hub genes and 5 DEmiRs (miR-221, miR-222, miR-506, miR-507, miR-876) were significantly associated with the overall survival probability. Conclusion: This bioinformatic study revealed the functions of several lncRNAs and their associated ceRNA network in UM metastasis. It provides a novel in silicon evidence for future experimental study on the pathogenesis of systemic metastasis in uveal melanoma, especially from the perspective of non-coding RNA.