胰腺癌转移相关基因的生物信息学分析

Objective: To provide theoretical basis for the molecular mechanism of the development of metastatic pancreatic cancer by screening differentially expressed genes based on bioinformatical analysis. Methods: We analyzed metastasis-related pancreatic cancer microarray datasets derived from GEO database. After preprocessing the data, R language package limma was used to screen differentially expressed genes. Then DAVID online tool was used for GO annotation and KEGG pathway enrichment analysis. STRING online database and Cytoscape software were utilized for protein-protein interaction network analysis. GEPIA online tool was used to evaluate prognostic performance. Results: We found 109 differentially expressed genes between metastatic pancreatic cancer tissues and primary pancreatic cancer tissues. Of them 49 genes were up-regulated and 60 were down-regulated. Enrichment analysis indicated that most of the genes were enriched in acute inflammatory response, complement and coagulation cascades, PPAR signaling pathway, and PI3K-Akt signaling pathway. Protein-protein interaction network analysis screened 2 key modules and 10 key genes (ORM1, IGFBP1, HPX, F2, APOA1, ALB, PLG, SERPINC1, KNG1 and INS). Prognostic analysis showed that 4 genes (SCG5, CRYBA2, CPE, and CHGB) were significantly associated with the patients' OS. Conclusion: The internal biological information in metastatic pancreatic cancer can be revealed by integrative bioinformaticalanalysis, providing theoretical basis for further research on molecular mechanism of metastatic pancreatic cancer.