Network-based study for identifying potentialbiomarkers for oral cancer diagnosis

Introduction: Oral cancer is the sixth most prevalent cancerworldwide. Until now, the molecular causes and genetic factorabnormalities are unclear. this study, analyzed submitted microarraydata with bioinformatic online tools to evaluate hubgenes and their potential in the prediction of the occurrence andrecurrence of oral cancer.Methods: Three microarray datasets including GSE۷۴۳۵۰,GSE۳۷۹۹۱ and GSE۳۱۰۵۶ were extracted from GEO datasetsin NCBI. In each dataset, differentially expressed genes (DEG)with p-value < ۰.۰۵ and logFc ≥ ± ۱.۵ were extracted. commonDEGs between the three datasets were identified by VennDiagrampackage using the R program. Gene ontology (GO) andthe Reactome enrichment analysis were done with the Enrichrtool. PPI network visualized by STRING database and hubgenes selected by the betweenness, degree and Maximal CliqueCentrality (MCC) algorithms in Cystoscope software via cytohubbaplugin. Selected hub genes validated by GEPIA onlineserver, performing overall survival (OS) and disease-free survival(DFS) analysis.Results: Reactome analysis shows that common DEGs are associatedwith the Assembly of Collagen Fibrils, Collagen Formationand degradation. GO term shows the biological processand molecular function of common DEGs in collagen-activatedtyrosine kinase receptor signaling pathway, lipid oxidation andCXCR۳/R chemokine receptor binding. By using Cytoscape wechoose ۱۰ top hub genes using ۳ different algorithms. Analysisof hub genes with GEPIA confirmed the main role of SMPX,IDO۱, PLAU, SPP۱, MMP۱ in oral cancer.Conclusion: Through Integrated bioinformatic and systems biologystudies, we proposed ۵ genes with the potential as diagnosticand or monitoring biomarkers in oral cancer.

کلیدواژه ها:

oral cancer ، PPI network ، hub genes ، biomarker