Risk assessment model constructed by differentially expressed lncRNAs for the prognosis of glioma

A risk assessment model was constructed using differentially expressed long non-coding (lnc)RNAs for the prognosis of glioma. Transcriptome sequencing of the lncRNAs and mRNAs from glioma samples were obtained from the TCGA database. The samples were divided into bad and good prognosis groups based on survival time, then differently expressed lncRNAs between these two groups were screened using DEseq and edgeR packages. Multivariate Cox regression analysis was performed to establish a risk assessment system according to the weighted regression coefficient of lncRNA expression. Survival analysis and receiver operating characteristic curve were conducted for the risk assessment model. Furthermore, the co-expression network of the screened lncRNAs was constructed, followed by the functional enrichment analysis for associated genes. A total of 117 lncRNAs were screened using edgeR and DEseq packages. Among all differently expressed lncRNAs, five lncRNAs (RP3-503A6, LINC00940, RP11-453M23, AC009411 and CDRT7) were identified to establish the risk assessment model. The risk assessment model demonstrated a good prognostic function with high area under the curve values in the training, validation and entire sets. The risk score was certified as an independent prognostic factor for gliomas. Multiple genes were screened to be co-expressed with these five lncRNAs. Functional enrichment analysis demonstrated that they were involved in cytoskeleton, adhesion and Janus kinase/signal transducer and activator of transcription signaling pathway-associated processes. The present study established a risk assessment model integrating five significantly different expressed lncRNAs, which may help to assess the prognosis of patients with glioma with increased accuracy.