Abstract:

Objective To screen the survival-associated alternative splicing （AS） events in the lung adenocarcinoma （LUAD） patients using bioinformatics method and construct the splicing factor （SF）-AS regulatory network， and to provide a new approach for the prognostic evaluation of the LUAD patients. Methods The transcriptome data and clinical data of the LUAD patients were downloaded from The Cancer Genome Atlas （TCGA） database， and the AS data of the LUAD patients were downloaded from the SF expression data and RNA target motifs（SpliceAid2）database. The survival-associated AS events of the LUAD patients were analyzed by univariate Cox regression analysis. The prognostic risk model of the LUAD patients was established by LASSO regression analysis and multivariate Cox regression analysis，and the risk score of each AS event was calculated based on the model. Kaplan-Meier （K-M） survival analysis and receiver operating characteristic （ROC） curve were used to evaluate the reliability of the model. The relationships between risk model or clinical parameters and independent prognosis of the LUAD patients were analyzed by using Cox regression analysis. The Pearson test was used to analyze the correlation of SF with AS events associated with prognosis. Cytoscape software was utilized to construct the SF-AS interaction network. Results A total of 43 948 cases of survival-associated AS events were screened， including 7 types of events. Exon skip （ES） events and alternate terminator （AT） events were the main events. The prognostic risk model of AS events was constructed， and the LUAD patients were divided into high and low risk groups based on the risk scores of the model.The K-M survival analysis results showed that the overall survival （OS） rate of the LUAD patients in high risk group was lower than that in low risk group （P<0.05）.The ROC curve analysis indicated that prognostic risk model of the LUAD patients had a good predictive effect，and the area under curve （AUC） was 0.824. The SF-AS regulatory network showed that 12 prognosis-related SFs modulated AS events positively or negatively，and they could predict the poor prognosis of the LUAD patients. Conclusion The prognostis-related AS events of LUAD patients and their upstream regulatory factor SF are screened，the SF-AS network is constructed，and the study provides the theoretical basis for further research on the correlation of AS events with prognosis of the LUAD patients.

Key words: Bioinformatics analysis, Lung adenocarcinoma, Alternative splicing, Splicing factor, The Cancer Genome Atlas database