Abstract
BACKGROUND: Alternative splicing (AS) acts on many tumors and its relationship with cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC) needs to be researched. METHODS: RNA sequencing data and clinical information of CESC cohorts were obtained from the Cancer Genome Atlas (TCGA) and SpliceSeq was used to analyze the splicing profile of mRNA in CESC. UpSetR displayed the intersections among AS events and univariate analysis chose survival-associated AS and splicing factor (SF) genes. Functional analysis was operated on Enrichr, STRING database and MCODE analysis were used to evaluate protein-protein interaction (PPI) information. LASSO and multivariate analysis constructed prognostic model and risk analysis of tumor infiltrating immune cells was also conducted. RESULTS: A total of 402 AS-generated genes were found to be associated with CESC prognosis. Functional analysis showed that Golgi to lysosome transport was enriched. PPI network suggested that UBA52 was most functional. Dendritic cells activated, dendritic cells resting, macrophages M0, mast cells resting, T cells CD4 memory activated and T cells CD8 were most correlative with the risk score. CONCLUSION: SFs and AS events can directly or indirectly affect the prognosis of CESC patients and this study identified SNRPA and CELF2 as two CESC-engaged SFs.