Premium
Identification of five long noncoding RNAs signature and risk score for prognosis of bladder urothelial carcinoma
Author(s) -
Zhang Chuanjie,
Li Zongtai,
Hu Jiateng,
Qi Feng,
Li Xiao,
Luo Jun
Publication year - 2020
Publication title -
journal of cellular biochemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.028
H-Index - 165
eISSN - 1097-4644
pISSN - 0730-2312
DOI - 10.1002/jcb.29330
Subject(s) - nomogram , receiver operating characteristic , univariate , proportional hazards model , oncology , bladder cancer , long non coding rna , multivariate statistics , lasso (programming language) , medicine , biology , cancer , gene , statistics , rna , computer science , mathematics , genetics , world wide web
Nowadays, an increasing number of studies illustrated that bladder urothelial cancer (BLCA) may act as the most common subtype of urological malignancies with a high rate of recurrence and metastasis. In this study, we attempted to establish a prognostic model and identify the possible pathway crosstalk. Long noncoding RNAs (lncRNAs) and mRNA expression and corresponding clinical information of patients with BLCA were downloaded from The Cancer Genome Atlas (TCGA). The differentially expressed genes analysis, univariate Cox analysis, the least absolute shrinkage, and selection operator Cox (LASSO Cox) regression model were then applied to identify five crucial lncRNAs (AC092725.1, AC104071.1, AL023584.1, AL132642.1, and AL137804.1). The multivariate cox analysis was utilized to calculate the regression coefficients (β i ). The risk‐score model was subsequently constructed as follows: (0.13541AC092725.1) + (0.20968AC104071.1) + (0.1525AL023584.1) − (0.14768AL132642.1) + (0.14387AL137804.1). Nomogram and assessment of overall survival (OS) prediction were verificated by the receiver operating characteristic curve in the testing group. As to 3‐, 5‐year OS prediction, the area under curve (AUC) for the nomogram of training data set was 0.83 and 0.86. Besides, the AUC (0.883 and 0.879) presented excellent predictive power in the testing group. In addition, the calibration plots validated the predictive performance of the nomogram. Weighted correlation network analysis (WGCNA) coupled with functional enrichment analysis contributed to explore the potential pathways, including PI3K‐Akt, HIF‐1, and Jak‐STAT signaling pathways. Construction of the risk‐score model and data analysis were both derived from multiple packages on the basis of the R platform chiefly.