| Objective:Globally,lung cancer is identified as the most frequently occurring malignant tumor with the highest mortality.Among them,85%of these patients can be classified as non-small cell lung cancer(NSCLC).In the past ten years,a large variety of targeted drugs have been marketed one after another,but why do patients' benefits differ greatly when they are targeted at the same target?Recently,researchers have found that alternative splicing is common in tumors.Alternative splicing is fundamental to transcriptome and proteome richness,and data from recently studies suggested a critical association between alternative splicing and oncogenic processes.At present,there are little related researches on alternative splicing in NSCLC.Therefore,in order to find its impact on tumor prognosis and the mechanism of action,we conducted this study.Methods:RNA transcriptome profiling,sequence data and clinic data of LUAD and LUSC group were retrieved from The Cancer Genome Atlasl(https://tcga-data.nci.nih.gov/).Besides,real-time quantitative PCR was conducted to validate the selected OS-associated DEAS events.Firstly,we screen TCGA cases by setting strict criteria and divide LUAD and LUSC into two subgroups according to gender.Process sequencing data through TCGA SpliceSeq software to obtain the patient's PSI.Then,the data was matched with clinical and expression data.Secondly,the paired samples of each group were screened out for further calculation.Statistical analyses are used to calculate which events are independent prognostic factors and establish a multivariate risk ratio regression model.Thirdly,the difference analysis of splicing factor was conducted in four groups(with adj.p<0.05).Then the correlation between differential SF and prognostic AS was calculated;Next,we selected the corresponding parent genes of OS-associated DEAS events as candidates for GO and KEGG pathway enrichment analysis.Fourth,the prognostic models along with clinicopathological variables described above were sent for univariate Cox analysis,and the significant results were further sent to develop a nomogram to estimate individual survival probability of patients.Fifth,Real-time Quantitative PCR was conducted to validate the selected OS-associated DEAS events.Results:Finally,eight AS events were identified as independent prognostic factors in the male and female subgroups in LUAD;seven types of AS events were determined as independent prognostic factors in the male and female subgroups in LUSC.The final composite models exhibited strong predictive power,and the AUCs of each group were all over 0.75 from 1 to 5 years.Besides,correlation network between SF and AS regulation was established,and the downstream pathway was enriched;In addition,the C indices of the Nomograms model are all above 0.7,and the correction curve also shows that the model is stable.Finally,four AS events were demonstrated to be differentially expressed in tumor tissue and normal tissue.Conclusion:In summary,this study performed a systematic analysis on prognostic splice events in NSCLC from gender and subtype perspectives.Then,we explored the upstream regulatory factors and downstream regulatory mechanisms of OS-related AS events that found in our study.More importantly,we constructed a well-executed nomogram that combines clinicopathological variables with four composite models.The final AS events obtained through layer screening may play an important role in tumorigenesis and deserve further study as molecular diagnostic biomarkers and therapeutic targets. |
PDF Full Text Request