ABSTRACT
Objective:To screen out and analyze the key genes of follicular lymphoma (FL) according to transcriptome data in Gene Expression Omnibus (GEO) database.Methods:Transcriptome datasets GSE32018 and GSE55267 were collected by using GEO database. R software was used for variance analysis to screen differentially expressed genes. FunRich 3.13 software was used to analyze common differential genes. The biological processes and pathways were analyzed with Cytoscape 3.7.2 software to screen potential genes related to the pathogenesis of FL. By analyzing clinical data from Oncomine database for survival analysis, the screened differential genes were verified.Results:A total of 141 up-regulated genes and 199 down-regulated genes were identified by differentially analyzing GSE32018 and GSE55267 datasets. Finally, 12 key genes including CXCL8, KRT19, CYCS, CDKN3, SFN, RRM2, FN1, APOE, CXCL12, VWF, GATA3 and TIMP1 were screened out; CYCS, CXCL8 and CXCL12 were closely related with the early survival rate of patients. Overexpression of CXCL12 and low expression of CYCS were found to be associated with poor prognosis of FL patients. CXCL8 expression was decreased in lymphoma tissues, but the relatively high expression of CXCL8 in survival analysis showed shortened overall survival, which might be related to the early development of FL. GO, KEGG and Reactome pathways were screened out including GO: 0001892, KEGG: 04115, R-HSA: 2559582, GO: 0060968, R-HSA: 6785807, GO: 0043627, GO: 0001936 and GO: 0043062.Conclusion:The selected genes CYCS, CXCL8 and CXCL12 may provide more effective biomarkers for the treatment of FL.