Font Size: a A A

Bioinformatics Analysis For Glioma Based On Microarray And Text Mining Technology

Posted on:2016-10-19Degree:DoctorType:Dissertation
Country:ChinaCandidate:B WeiFull Text:PDF
GTID:1224330467995389Subject:Surgery
Abstract/Summary:PDF Full Text Request
Objective:Glioma is now a common tumor disease, and which is considered to be one ofthe most malignant human neoplaamas. Although several genes have been indentifiedto be associated with tumorigenesis and anaplastic progression of gliomas, theircontribution to the molecular classification of astrocytic has been limited.In this study, bioinformatic analysis of glioblastomas was performed usingmicroarray analysis and text mining to screen differentially expressed genes(DEGs),differentially co-expressed genes (DCGs) and differentially co-expressed links(DCLs).Then, protein-protein interaction (PPI) network, regulatory network, andmolecular interaction network of biology pathways were constructed, respectively.Finally, enrichment analysis was also performed. This study was aimed to clarify themechanism ofglioblastomasdevelopment, and to provide basis for glioblastomasdiagnosis and treatment.Methods:Gene expression profile (GSE4290), including77glioblastomassamples and23normal control samples, was downloaded from Gene Expression Omnibusdatabase.DEGs were screened using Limma package of R software, and theP-value<0.05and|logFC|>2were chosen as cut-off criterion. GO enrichment analysiswas performed using DAVID. PPI network and sub-network were respectivelyconstructed by Cytoscape software and ClusterONE plug-in ofCytoscape software.Then, functional enrichments of sub-networks were performed.Firstly, DEGs were screened using Affy and Limma package of R software, andthe P-value<0.05and|logFC|>0.6were chosen as cut-off criterion. Then, DCGs and DCLs were identified by DCe、DCp and DCsum function of DCGL package in Rsoftware, and the q <0.25was considered as cut-off criterion. KEGG pathwayenrichment analysis was performed for DCGs using DAVID. Regulatory network andsub-network of the top5modules were respectively constructed by Cytoscapesoftware and ClusterONE plug-in ofCytoscape software.Based on text mining analysis and MGD database, the mutation genes related toglioblastomaswere screened. Then, molecular interaction networkwas constructedusing Cytoscape software, and KEGG pathway and GO functional enrichmentanalysis were performed by DAVID.Results:A total of548DEGs were identified, including441down-and107up-regulatedgenes. In PPI network, there were1305nodes and1604edges. The hub genes in thetop5significant sub-networks were FN1, GNAO1, STX1A, CDK1and CHGB,respectively. The significant enrichment GO terms were extracellular region part,vesicle-mediated transport, cell cycle process, and nucleoplasm. The significantenrichment KEGG pathways were ECM-receptor interaction, long-term depression,SNARE interactions in vesicular, and cell cycle.In total,999DCGs and1833200DCLs were obtained. In regulatory network,1441nodes and2127edges were found. The hub genes in the top5significantsub-networks were STAT1, STAT3, STAT4, PLAU, EPAS1, BPTF, CBFB, NFYB,and EGR3. The significant enrichment GO terms were enzyme binding, proteinmodification by small protein conjugation or removal, muscle cell differentiation,mitotic cell cycle, and pore complex.A total of52518molecular interaction pairs were screened, of which27526werethe homologous interaction pairs between human and mouse. Nine mutation genesrelated to glioblastomaswere screened. In molecular interaction network,875nodesand1068edges were identified, and the genes with higher degree were TP53,CDKN2A, PTEN, NF1, and TG. The significant enrichment GO terms werenucleoplasm and regulation of cell death. The significant enrichment KEGG pathways were p53signaling pathway, pathways in cancer, and cell cycle.Conclusions:DEGs andDCGs,involving in glioblastomas development, were screened.PPInetwork and regulatory network, and molecular interaction network of biologypathways were constructed, respectively. DEGs,DCGs, mutation genes, andsignificant enrichments GO terms and KEGG pathaways might contribute to themechanism ofglioblastomas development. There results could be used for the drugscreening and drug designing for glioblastomas treatment.
Keywords/Search Tags:DNA chip, text mining, glioma, biological pathway
PDF Full Text Request
Related items