Role of survival-associated alternative splicing events in the prognosis of ovarian cancer

Abstract

Objective Alternative splicing (AS) events play a crucial role in the tumorigenesis and progression of various cancers. In the present study, we aimed to identify specific AS events, which might be prognostic markers and therapeutic targets for ovarian cancer (OV). Methods Transcriptome data, clinical information, and Percent Spliced In (PSI) values were downloaded from TCGA database and TCGA SpliceSeq to explore the role of AS events in the prognosis of OV patients. Univariate and multivariate Cox regression analyses were performed to identify survival-associated AS events and develop multi-AS-based prognostic models. The K-M curves and ROC curves were conducted based on prognostic AS event models. Moreover, a splicing regulatory network was established according to the correlation between AS events and splicing factors (SFs). Finally, we performed functional enrichment analysis by GO terms and KEGG pathways. Results We identified 1,472 AS events that were associated with the survival of OV patients, and exon skipping (ES) was the most important type. We also found that prognostic models based on AS events were good predictors of OV prognosis, which could discriminate the high-risk group from the low-risk group (P < 0.05). Notably, the AUC value of AD, AP, AT, ES, ME, and the whole cohort was more than 0.70, indicating that these six models had valuable prediction strength. The risk score of prognostic models was identified as an independent prognostic factor. Furthermore, the AS-SF correlation network revealed several hub SF genes, including DDX39B, PNN, LUC7L3, ZC3H4 and SRSF11, and so on. Conclusions In the present study, we constructed powerful prognostic predictors for OV patients and uncovered interesting splicing networks. Collectively, our findings provided valuable insights into the underlying mechanisms of OV.