[Exploring of a prognostic long non-coding RNA signature of hepatocellular carcinoma by using public database].

• DOI: 10.3760/cma.j.issn.0254-6450.2019.07.014
• Publication Date: 2019-07-10
• Authors: Junhui Xiao; K Wang; Y Liu; Z W Li; Yijun Zhou; H Z Wang; Junyang Lu; S S Cheng; Shaobin Wei
• Keywords: Quartile

Objective: To explore an effective long non-coding RNA (lncRNA) signature in predicting the prognosis of hepatocellular carcinoma through analysis on sequencing data patients and peritumoral tissues Cancer Genome Atlas (TCGA) database. Methods: The clinical characteristics 377 were obtained from TCGA database by end February 2018. Then, differentially expressed lncRNAs between 50 pairs tumor explored using student's t-test. Next, a lncRNA was established LASSO Cox regression analysis. All divided into four groups (<P(25), P(25)-, P(50)-, ≥P(75)) based cut-off quartiles signature. Finally, compared with control group (<P(25)), hazard ratios (HRs) three (P(25)-, calculated regression. survival outcomes to evaluate capacity model. Results: A total 951 identified tissues. three-lncRNA signature, including LNCSRLR, MKLN1-AS ZFPM2-AS1, predict patients. outcome suggested that death risk ≥P(75) 1.57 times larger than <P(25) (95%CI: 1.06-2.31, P<0.05). Conclusion: which significantly associated data.目的： 通过对癌症和肿瘤基因图谱（TCGA）公共数据库中肝细胞癌病例癌和癌旁组织RNA测序数据的分析，挖掘与肝细胞癌预后相关的长链非编码RNA（lncRNA）分子标签。 方法： 截至2018年2月，从TCGA数据库中获得377例肝细胞癌病例的癌及癌旁组织RNAseq数据及临床预后信息，将50对癌和癌旁组织的lncRNA表达水平进行差异t检验分析，进而采用LASSO Cox回归分析筛选肝细胞癌预后相关的lncRNA，并构建lncRNA分子标签。将所有病例按分子标签表达水平分为4组（<P(25)、P(25)~、P(50)~、≥P(75)），采用Cox回归计算P(25)~、P(50)~、≥P(75)组相对于<P(25)组的预后风险比，进而评估分子标签表达水平对肝细胞癌病例总体生存率的影响。 结果： 筛选出951个癌和癌旁组织中表达水平有统计学意义差异的lncRNA，通过LASSO Cox回归分析进一步筛选出3个lncRNA（LNCSRLR、MKLN1-AS及ZFPM2-AS1），并构建分子标签。分子标签表达水平≥P(75)组的死亡风险是<P(25)组的1.57倍（95%CI：1.06~2.31，P<0.05）。 结论： 通过对TCGA数据库的挖掘，由LNCSRLR、MKLN1-AS及ZFPM2-AS1构建的lncRNA分子标签表达水平与肝细胞癌病例的预后有关。.