基于 TCGA 数据库和生物信息学的自噬相关长非编码 RNA 风险签名预测肾透明细胞癌预后
A Risk Signature with Autophagy-Related Long Noncoding RNAs for Predicting the Prognosis of Clear Cell Renal Cell Carcinoma: Based on the TCGA Database and Bioinformatics.
机构信息
Medical School of Chinese PLA, Beijing 100853, China.
Department of Urology, The Third Medical Centre, Chinese PLA (People's Liberation Army) General Hospital, Beijing 100853, China.
出版信息
Dis Markers. 2021 May 7;2021:8849977. doi: 10.1155/2021/8849977. eCollection 2021.
BACKGROUND
Disorders of autophagic processes have been reported to affect the survival outcome of clear cell renal cell carcinoma (ccRCC) patients. The purpose of our study was to identify and validate the candidate prognostic long noncoding RNA signature of autophagy.
METHODS
Transcriptome profiles were obtained from The Cancer Genome Atlas. The autophagy gene list was obtained from the Human Autophagy Database. Based on coexpression analysis, we obtained a list of autophagy-related lncRNAs (ARlncRNAs). GO enrichment analysis and KEGG pathway analysis were conducted to explore the functional annotation of these ARlncRNAs. Univariate and multivariate Cox regression analyses were conducted to elucidate the correlation between overall survival and the expression level of each ARlncRNAs. We then established a prognostic signature that was a linear combination of the regression coefficients from the multivariate Cox regression model () multiplied by the expression levels of the respective ARlncRNAs in the training cohort. The predictive performance was tested in the validation cohort. Additionally, the independence of the risk signature was assessed, and the relationship between the risk signature and conventional clinicopathological features was explored.
RESULTS
Seven autophagy-related lncRNAs with prognostic value (SNHG3, SNHG17, MELTF-AS1, HOTAIRM1, EPB41L4A-DT, AP003352.1, and AC145423.2) were identified and integrated into a risk signature, dividing patients into low-risk and high-risk groups. The risk signature was independent of conventional clinical characteristics as a prognostic indicator of ccRCC (HR, 1.074, 95% confidence interval: 1.036-1.113, < 0.001) and was valuable in the prediction of ccRCC progression.
CONCLUSION
Our risk signature has potential prognostic value in ccRCC, and these ARlncRNAs may play a significant role in ccRCC tumor biology.
背景
自噬过程的紊乱被报道会影响透明细胞肾细胞癌（ccRCC）患者的生存结果。本研究的目的是鉴定和验证自噬的候选预后长非编码 RNA 特征。
方法
从癌症基因组图谱（The Cancer Genome Atlas）获取转录组谱。从人类自噬数据库（Human Autophagy Database）获取自噬基因列表。基于共表达分析，我们获得了一组自噬相关长非编码 RNA（ARlncRNAs）。进行 GO 富集分析和 KEGG 通路分析，以探讨这些 ARlncRNAs 的功能注释。进行单变量和多变量 Cox 回归分析，以阐明总生存期与每个 ARlncRNAs 的表达水平之间的相关性。然后，我们在训练队列中建立了一个预后特征，它是多元 Cox 回归模型的回归系数乘以各自 ARlncRNAs 在训练队列中的表达水平的线性组合。在验证队列中测试预测性能。此外，评估了风险特征的独立性，并探讨了风险特征与常规临床病理特征之间的关系。
结果
确定了 7 个具有预后价值的自噬相关 lncRNAs（SNHG3、SNHG17、MELTF-AS1、HOTAIRM1、EPB41L4A-DT、AP003352.1 和 AC145423.2），并将其整合到一个风险特征中，将患者分为低风险和高风险组。该风险特征作为 ccRCC 的预后指标独立于常规临床特征（HR，1.074，95%置信区间：1.036-1.113，<0.001），并且对 ccRCC 进展具有预测价值。
结论
我们的风险特征在 ccRCC 中具有潜在的预后价值，这些 ARlncRNAs 可能在 ccRCC 肿瘤生物学中发挥重要作用。
Zotero 插件