Rong-Quan He1,2,3,4,5, Xian-Guo Zhou1,2,3,4,5, Qiao-Yong Yi1,2,3,4,5, Cai-Wang Deng1,2,3,4,5, Jia-Min Gao1,2,3,4,5, Gang Chen2,3,4,5,6, Qiu-Yan Wang1,2,3,4,5. 1. Department of Biochemistry and Molecular Biology, Guangxi Medical University, Nanning, China. 2. Center for Genomic and Personalized Medicine, Guangxi Medical University, Nanning, China. 3. Guangxi Collaborative Innovation Center for Genomic and Personalized Medicine, Nanning, China. 4. Guangxi Key Laboratory for Genomic and Personalized Medicine, Guangxi Collaborative Innovation Center for Genomic and Personalized Medicine, Nanning, China. 5. Guangxi Collaborative Innovation Center for Biomedicine, Guangxi Medical University, Nanning, China. 6. Department of Pathology, First Affiliated Hospital of Guangxi Medical University, Nanning, China.
Abstract
BACKGROUND/AIMS: Increasing evidences indicated the important roles of alternative splicing in the progression and prognosis of bladder urothelial carcinoma (BLCA). However, most previous research has focused on one or several alternative splicing events, without a comprehensive evaluation of the prognostic value of splicing events in BLCA. In this study, we aimed to determine risk scores for predicting prognosis of BLCA patients based on splicing events. METHODS: RNA-sequencing data and clinical information of BLCA patients were downloaded from The Cancer Genome Atlas, and data of splicing events were obtained from the SpliceSeq database. Univariate and multivariate Cox regression analyses were employed to identify survival-associated alternative spicing events (SASEs) and to calculate risk scores. Protein-protein interaction analysis of genes of the SASEs was performed using STRING, a database of known and predicted protein-protein interactions, and pathway enrichment analysis of the genes was implemented using the Database for Annotation, Visualization and Integrated Discovery (version 6.8). Receiver operating characteristic (ROC) curves and Kaplan-Meier analysis were used to evaluate the clinical significance of genes from the SASEs for building a risk score in BLCA. Correlation between splicing events of splicing factors and non-splicing factors were analyzed with Pearson correlation coefficient. A potential regulatory network was then built using Cytoscape 3.5. RESULTS: In total, 39,508 alternative splicing events in 317 patients with BLCA were analyzed, including 4,632 SASEs. The area under the curve of the ROC of risk score (all) was 0.748 for predicting survival status of BLCA patients. Low- and high-risk score groups classified using the median "risk score (all)" value displayed remarkably different survival time (Low vs. High = 3304.841±239.758 vs 1198.614±152.460 days). The potential regulatory network with SASEs of splicing factors and other genes was constructed, which might be part of the biological mechanisms associated with prognosis of BLCA patients. CONCLUSIONS: In this study, prognostic signatures constructed using splicing events could be used for predicting the prognosis of BLCA patients.
BACKGROUND/AIMS: Increasing evidences indicated the important roles of alternative splicing in the progression and prognosis of bladder urothelial carcinoma (BLCA). However, most previous research has focused on one or several alternative splicing events, without a comprehensive evaluation of the prognostic value of splicing events in BLCA. In this study, we aimed to determine risk scores for predicting prognosis of BLCA patients based on splicing events. METHODS: RNA-sequencing data and clinical information of BLCA patients were downloaded from The Cancer Genome Atlas, and data of splicing events were obtained from the SpliceSeq database. Univariate and multivariate Cox regression analyses were employed to identify survival-associated alternative spicing events (SASEs) and to calculate risk scores. Protein-protein interaction analysis of genes of the SASEs was performed using STRING, a database of known and predicted protein-protein interactions, and pathway enrichment analysis of the genes was implemented using the Database for Annotation, Visualization and Integrated Discovery (version 6.8). Receiver operating characteristic (ROC) curves and Kaplan-Meier analysis were used to evaluate the clinical significance of genes from the SASEs for building a risk score in BLCA. Correlation between splicing events of splicing factors and non-splicing factors were analyzed with Pearson correlation coefficient. A potential regulatory network was then built using Cytoscape 3.5. RESULTS: In total, 39,508 alternative splicing events in 317 patients with BLCA were analyzed, including 4,632 SASEs. The area under the curve of the ROC of risk score (all) was 0.748 for predicting survival status of BLCA patients. Low- and high-risk score groups classified using the median "risk score (all)" value displayed remarkably different survival time (Low vs. High = 3304.841±239.758 vs 1198.614±152.460 days). The potential regulatory network with SASEs of splicing factors and other genes was constructed, which might be part of the biological mechanisms associated with prognosis of BLCA patients. CONCLUSIONS: In this study, prognostic signatures constructed using splicing events could be used for predicting the prognosis of BLCA patients.