An alternative splicing signature model for predicting hepatocellular carcinoma-specific survival.

Abstract

Alternative splicing (AS) is a transcriptional regulation mechanism, which can expand the coding ability of genome and contribute to the occurrence and development of cancer. A systematic analysis of AS in hepatocellular carcinoma (HCC) is lacking and urgently needed. Univariate and multivariate Cox regression analyses were used to distinguish survival-related AS events and to calculate the risk score. Kaplan-Meier analysis and receiver operating characteristic (ROC) curves were used to evaluate the AS events' clinical significance to build a risk model in HCC. Data of AS events was obtained from the Splice-Seq database. The corresponding clinical information of HCC was downloaded from The Cancer Genome Atlas (TCGA) data portal. We analyzed 78,878 AS events from 13,045 genes in HCC patients. A total of 2,440 and 2,888 AS events were significantly related to HCC patients' disease-free survival (DFS) and overall survival (OS). The two prognostic models (DFS and OS) were constructed based on a total of seven AS types from survival-related AS events above. The area under the curve (AUC) of the ROC curves was 0.769 in the DFS cohort and 0.886 in the OS cohort. The prognostic model constructed by AS events can be used to predict the prognosis of HCC patients and provide potential therapeutic targets for further validation.