Identification Of Hub Gene GRIN1 Correlated With Histological Grade And Prognosis Of Glioma By Weighted Gene Co-Expression Network Analysis

Objective: Gliomas originate from neuroepithelial tissue and have an incidence rate of approximately 8/100,000 adults per year,making them the most common primary intracranial tumor.Even with the most aggressive clinical treatment regimen,the median survival of glioblastoma patients is only 12-15 months.The lack of satisfactory efficacy of conventional surgical treatment,radiotherapy,and chemotherapy has given rise to many studies on glioma-targeted markers.Bioinformatics is the study of biological information acquisition,processing,storage,dissemination,analysis,and interpretation of all aspects of the discipline,but also with the rapid development of life sciences and computer science,life sciences and computer science combined to form a new discipline.The integrated use of biology,computer science,and information technology reveals the biological mysteries endowed with large and complex biological data.The aim of this study is to identify core genes associated with glioma pathological classification and prognosis by means of weighted gene network co-expression analysis in bioinformatics.Methods: 1.Differentially expressed genes highly associated with glioma were presented in both the GEO dataset and TCGA dataset,with 511 differential genes identified from the GEO dataset and 5331 differential genes identified from the TCGA dataset;the turquoise module in the GEO dataset(r =-0.73,p = 9e-31)and the TCGA dataset(r =-0.24,p = 3e-04)in the turquoise module had the most significant negative correlation regulation with glioma.2.The clinical data and transcriptomic data obtained from glioma patients were processed by R language programming software(version 3.6.3)and PERL programming software;the differential genes in gliomas in the GEO and TCGA datasets were found by loading packages such as Limma,edge R,pheatmap,ggplot2 and WGCNA,and the The differential genes were subjected to weighted gene network co-expression analysis.The differential gene clusters and weighted gene clusters in gliomas were visualized by constructing module maps,tree maps,heat maps and volcano maps.3.Venn diagram was drawn by R language programming software and Venn Diagram software engineering package to construct intersecting genes of glioma differential gene clusters.4.The biological functions of the intersecting gene clusters were analyzed by R programming software and software engineering packages such as colorspace,stringi,ggplot2,Bioc Manager,Dose,cluster Profiler and enrichplot for the Kyoto Encyclopedia of Genes and Genomes(KEGG)and Gene Ontology(GO)bio functional enrichment analysis.5.Based on STRING,an online database,we established the protein interaction network of intersecting genes based on the evidence of text mining,previous experimental reports,the information in databases,and evidence of gene co-expression,co-evolution,and co-fusion.6.Topological analysis of protein interaction networks of intersecting genes was performed by bioinformatics software Cytoscape(version3.8.2)and plug-in Cytohubba;10 hub genes strongly associated with glioma were screened from intersecting genes based on Maximum centrality criterion(MCC).7.Based on R language programming software and Survival and Survminer,we performed prognostic correlation analysis of 10 core genes using Kaplan-Meier’s statistical method;meanwhile,we performed the multi-subgroup analysis of expression and prognosis of 10 core genes using online open databases GEPIA and LOGpc.8.Clinical correlation analysis of GRIN1,the hub gene with the most significant expression differences and strongest prognostic correlation in glioma,based on the clinical trait files of glioma patients in the TCGA database,GRIN1 correlation analysis was performed on glioma patients of different ages and genders using R language programming software and Limma and ggpubr software engineering packages.9.The open online database HPA checked the protein expression differences of GRIN1 in glioma and normal brain tissues;meanwhile,nine patients(three glioblastomas,three low-grade gliomas,and three cerebrovascular malformations)operated at the Department of Neurosurgery,Shengjing Hospital,China Medical University from April 30,2021,to August 27,2021,were subjected to immunohistochemical analysis to identify differences in protein expression of GRIN1 in different glioma grades and normal brain tissue.Results: 1.Differential genes highly associated with glioma were presented in both the GEO dataset and TCGA dataset,with 511 differential genes identified from the GEO dataset and 5331 differential genes identified from the TCGA dataset;the turquoise module in the GEO dataset(r =-0.73,p = 9e-31)and the TCGA dataset(r =-0.24,p = 3e-04)in the turquoise module had the most significant negative correlation regulation with glioma.2.Venn diagram yielded 185 intersecting genes in the GEO dataset differential gene cluster,TCGA dataset differential gene cluster,TCGA dataset turquoise module,and GEO dataset turquoise module.3.The KEGG enrichment analysis showed that 185 intersecting genes were enriched in the calcium signaling pathway;the GO enrichment analysis showed that in the Biological process module(BP),intersecting genes mainly played the role of regulating chemical synaptic transmission;in the cellular component module(CC),intersecting genes mainly played the role of pre-synaptic regulation;in the Molecular function module(MF),intersecting genes mainly played the role of regulating the activity of metal ion transmembrane transporter.4.The STRING database framed a protein interaction network of 185 intersecting genes(r = 0.4).Cytoscape and the plug-in Cytohubba performed an MCC-based topological analysis of this protein interaction network to obtain ten core genes highly associated with glioma: DLG4,GRIN1,GNB4,GNB5,GRIN2 B,VAMP2,DLG3,F2 R,ADRA1B,and CCK.5.By Kaplan-Meier calculation of survival differences,we found that only GRIN1 had a more significant prognostic value in glioma patients,and high expression of GRIN1 revealed a better prognosis;meanwhile,when searching the online databases GEPIA and LOGpc for multi-subgroup analysis;we found that GRIN1 in normal brain tissue and glioma tissue had a significant We also found a statistically significant difference between GRIN1 in normal brain tissue and glioma tissues when we searched the online databases GEPIA and LOGpc for multi-subgroup analysis and revealed that high expression state of GRIN1 resulted in a better prognosis for glioma patients.6.The correlation analysis of clinical traits of GRIN1 revealed no statistically significant differences in GRIN1 among glioma patients of different genders and age groups.7.By searching the online open database HPA,GRIN1 showed low expression in glioma tissues and high expression in normal tissues;while for the collected pathological sections,we found low expression of GRIN1 in high-grade and low-grade gliomas,while in normal brain tissues,GRIN1 was highly expressed.Conclusion:1.Bioinformatics-based co-expression analysis of weighted gene networks can identify core genes in gliomas and provide potential targets and biomarkers to guide prognosis for the treatment of gliomas.2.GRIN1 is lowly expressed in gliomas and highly expressed in normal brain tissues;a high expression state of GRIN1 predicts a more favorable long-term prognosis for glioma patients.3.GRIN1 could be a new target for the treatment of glioma and guide the prognosis of glioma patients.