Integrated Gene Expression Data-Driven Identification of Molecular Signatures, Prognostic Biomarkers, and Drug Targets for Glioblastoma

Glioblastoma (GBM) is a highly prevalent and deadly brain tumor with high mortality rates, especially among adults. Despite extensive research, the underlying mechanisms driving its progression remain poorly understood. Computational analysis offers a powerful approach to explore potential prognostic biomarkers, drug targets, and therapeutic agents for GBM. In this study, we utilized three gene expression datasets from the Gene Expression Omnibus (GEO) database to identify differentially expressed genes (DEGs) associated with GBM progression. Our goal was to uncover key molecular players implicated in GBM pathogenesis and potential avenues for targeted therapy. Analysis of the gene expression datasets revealed a total of 78 common DEGs that are potentially involved in GBM progression. Through further investigation, we identified nine hub DEGs that are highly interconnected in protein–protein interaction (PPI) networks, indicating their central role in GBM biology. Gene Ontology (GO) and pathway enrichment analyses provided insights into the biological processes and immunological pathways influenced by these DEGs. Among the nine identified DEGs, survival analysis demonstrated that increased expression of GMFG correlated with decreased patient survival rates in GBM, suggesting its potential as a prognostic biomarker and preventive target for GBM. Furthermore, molecular docking and ADMET analysis identified two compounds from the NIH clinical collection that showed promising interactions with the GMFG protein. Besides, a 100 nanosecond molecular dynamics (MD) simulation evaluated the conformational changes and the binding strength. Our study highlights the potential of GMFG as both a prognostic biomarker and a therapeutic target for GBM. The identification of GMFG and its associated pathways provides valuable insights into the molecular mechanisms driving GBM progression. Moreover, the identification of candidate compounds with potential interactions with GMFG offers exciting possibilities for targeted therapy development. However, further laboratory experiments are required to validate the role of GMFG in GBM pathogenesis and to assess the efficacy of potential therapeutic agents targeting this molecule.


Introduction
Glioblastoma (GBM) is a highly aggressive and deadly form of brain cancer that arises from astrocytes, the supportive cells of the brain.It is also referred to as GBM multiforme (GBM) [1].It is the most common primary brain tumor in adults and accounts for approximately 47% of all malignant brain tumors [2].The prognosis for GBM patients is extremely poor, with a survival of 15 to 23 months and a less than 6% chance of surviving beyond 5 years after diagnosis [3].The incidence of GBM is particularly prominent in North America, Australia, and northern and western Europe [4].It has a yearly incidence rate of 4.23 per 100,000 individuals [5].According to experimental data, GBM can develop into various types of tumors and contain a subset of highly tumourigenic cells known as GBM stem cells, which are believed to be responsible for recurrent GBM [6,7].GBMs can be classified as primary or secondary GBMs, as mentioned earlier [8,9].Primary GBM, which develops rapidly in older people (mean age 55) without prior brain malignancies, accounts for nearly 90% of cases and is characterized by mutations in IDH1, PDGFRA abnormalities, EGFR overexpression, PTEN mutation, and chromosome loss [10][11][12].In contrast, secondary GBM, which accounts for approximately 10% of cases, evolves from preexisting low-grade gliomas and is marked by mutations in IDH1, IDH2, MGMT, TP53, and alterations in 19q.Molecular markers, such as EGFR, PDGFRA, NF1, IDH1, MGMT, p53, and PTEN are utilized for characterizing GBM [13][14][15].
Among the molecular biomarkers identified in GBM, the IDH1, MGMT, and EGFR were demonstrated to have clinical significance [16][17][18].IDH1 mutations are typically associated with younger patients and secondary GBM, often accompanied by TP53 mutations, and are incorporated into WHO diagnostic guidelines as a positive prognostic indicator [16,19,20].MGMT promoter methylation, while not included in current WHO classifications, is a crucial prognostic marker that predicts response to alkylating agents like temozolomide [17,21].Promoter methylation-induced MGMT silencing enhances temozolomide's cytotoxic effects, potentially increasing patient survival [22].EGFR abnormalities, driven by amplification or the EGFRvIII mutation, are linked to higher tumor malignancy and poorer prognosis [18].EGFR serves both as a prognostic marker and a therapeutic target, with anti-EGFR therapies like Gefitinib and Erlotinib being tested in clinical trials to inhibit its downstream signaling by preventing tyrosine residue phosphorylation [23,24].While IDH1 mutations offer a better prognosis, they are less prevalent in primary GBM, limiting their applicability in a significant portion of patients [25].Additionally, although MGMT promoter methylation can forecast response to temozolomide, not all patients with methylated MGMT experience equal benefits, suggesting the influence of other factors on treatment response [26].Moreover, clinical trials of EGFR-targeted therapies have demonstrated limited efficacy, partly due to tumor heterogeneity and acquired resistance mechanisms [27].The intricate interplay between these biomarkers and other molecular alterations in GBM necessitates further investigation to fully understand their clinical relevance and therapeutic implications.Exploring more biomarkers beyond IDH1, MGMT, and EGFR could reveal new GBM subtypes, improving personalized treatment and patient outcomes.Integrating multiple biomarkers may enhance precision medicine, tailoring therapies to each tumor's unique profile [28].Biomarker discovery not only identifies new treatment targets but also guides the development of innovative therapies, making ongoing efforts essential for advancing GBM understanding and patient care.
For most GBM patients, no specific risk factors have been identified.The only known external environmental risk factor for glioma is exposure to ionizing radiation [29,30].Other factors, such as viral triggers (human cytomegalovirus) [31], adolescent obesity [32], and a family history of cancer [33], are still under investigation.Current research is focused on identifying germline polymorphisms associated with an increased risk of GBM, as genetic factors play a role in determining the level of risk associated with these exposures [30].The currently available treatments for GBM include maximal resection (complete resection is very rare, because these tumors spread throughout the body) followed by radiotherapy with concurrent adjuvant therapies, such as temozolomide (TMZ).Bevacizumab, which inhibits circulating vascular endothelial growth factor (VEGF), is often recommended for patients with progressive cancer and has recently been used in combination with lomustine (CCNU) [34].Despite numerous efforts, there has been no improvement in the survival rates of most GBM patients, and GBM patients suffer recurrence as a result of molecular heterogeneity and the challenge of drug penetration across the blood-brain barrier (BBB).However, recent developments in genomics and transcriptomics have led to the discovery of specific molecular signatures of GBM that enable us to better understand the molecular mechanism of GBM.Identifying the crucial molecular signatures, biomarkers, and therapeutic targets in GBM could significantly improve treatment strategies and decrease fatalities.A prognostic biomarker refers to a genetic marker that forecasts the probability of a forthcoming clinical event, recurrence of a disease, or its progression within a population.The biological features of biomarkers help predict the course of a disease or the response to treatment among patients.
The Gene Expression Omnibus (GEO) database serves as a global online repository for microarray gene expression data and contains a wide range of readily accessible functional genomics datasets.The exploration of microarray data can provide insights into the pathology and molecular mechanisms underlying GBM.Therefore, we analyzed microarray data using in silico techniques to examine the role of genes in pathogenic processes.The major objective of this study was to identify potential biomarkers, molecular signatures, and therapeutic agents that could contribute to the early detection of GBM and serve as molecular targets for drug candidates.

Materials and Methods
Publicly accessible data were analyzed by the following methods to meet the aim set for this study.To identify the significant components of the molecular pathways associated with GBM, we retrieved three gene expression datasets (Table 1) from the GEO database [35].The datasets were chosen, as they include samples from both GBM tissues and normal brain tissues, allowing for direct comparisons with high-quality data.The datasets focus specifically on GBM, excluding other types of brain tumors or unrelated diseases, ensuring relevance to the study.Each dataset contains a sufficient number of samples to ensure robust statistical power.The use of consistent platforms, namely, GPL570 and GPL8300, helps minimize technical variability and enhances the 3 BioMed Research International reliability of the findings.Details of the selected datasets and their DEGs are shown in Table 1.

Identification of Differentially Expressed Genes (DEGs).
The GEO 2R (https://www.ncbi.nlm.nih.gov/geo/geo2r/), an interactive web server for analyzing microarray datasets, was used to identify DEGs in the GSE50161, GSE12657, and GSE15824 datasets by comparing normal and disease samples.The datasets were analyzed and normalized by the Limma [36] package to identify DEGs in GBM.For all the datasets, a threshold of p value < 0.05 and logFC > 1 0 was set to determine significant DEGs.Common genes were also identified from the datasets via Venn analysis using the web tool jvenn [37].
2.3.Hierarchical Clustering.We utilized the Cluster 3.0 tool to conduct clustering analysis on the selected gene expression data and employed another tool called Java TreeView to visualize the resulting analysis.The clustering process started by creating a tab-delimited text file to use as input for the Cluster 3.0 tool.The initial step in hierarchical clustering involves calculating the distance matrix for the gene expression data.After this matrix of distances is computed, the clustering process can begin.The generated CDT file from Cluster 3.0 was the input file for Java TreeView, which produced a dendrogram displaying the hierarchical clustering of genes, including both gene tree and array tree.Hierarchical clustering methods arrange genes into a tree-like formation according to their similarity [38,39].Cluster 3.0 represents an enhanced iteration of the Cluster program, utilizing the C Clustering Library [40].
2.4.Construction of the PPI Network.Utilizing the STRING (https://string-db.org/)database [41], a protein-protein interaction (PPI) network of proteins encoded by common DEGs was constructed to illustrate how the defined DEGs and proteins physically and functionally interact with each other.We also identified hub genes from the network by using the MCODE algorithm [42].The generated network file was customized in Cytoscape (https://cytoscape.org)[43].

Gene Ontology (GO) and Pathway Enrichment Analysis.
Gene set enrichment analysis (GSEA) is a computational and statistical methodology that determines whether a collection of determined genes exhibits statistical significance under various biological conditions.Pathway annotations were acquired from the Reactome database.GO terms and pathways for the present study were obtained using the EnrichR (https://amp.pharm.mssm.edu/Enrichr/)platform [44].For all analyses, a p value < .05 was considered to indicate statistical significance.

Identification of Transcription Factor (TF) Interactions
With Hub DEGs.TFs interact with genes to control gene expression by activating or inactivating transcription.We identified TF-gene interactions with the hub DEGs using the JASPAR [45] and TRANSFAC [46] databases through EnrichR.

Survival Analysis.
To examine the prognostic performance of the hub genes in detecting GBM, a multivariate survival analysis of GBM patients was performed based on the expression of the hub genes by using the GEPIA2 web tool (http://gepia2.cancer-pku.cn/)[47].The significance level was set to a p value < 0.05.
2.8.Validation of Genes.The selected key DEG was the GMFG gene, which was validated through differential expression analysis via GEPIA2.The mRNA expression pattern of the GMFG gene in GBM tissue was determined using two servers, UALCAN (http://ualcan.path.uab.edu) and the OncoDB server (http://oncodb.org/),to increase the fidelity of the findings.The GEPIA 2 is an online platform that enables the analysis of RNA sequencing data from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) datasets on malignant and normal tissues [47].The UALCAN web portal enables the cancer research community to analyze and receive cancer transcriptome, proteome, 4 BioMed Research International and patient survival data [48].OncoDB also facilitates the investigation of differential gene expression in malignant tissues and the correlation of gene expression with the clinical outcome of cancer patients [49].The differential expression of GMFG in cancerous conditions was also investigated by analyzing the immunohistochemistry of GBM and healthy cells curated from the Human Protein Atlas database [50].
2.9.Structural Preparation of Proteins and Ligands.The crystal structure of human GMFG, PDB ID 3L50, was extracted from the Protein Data Bank database (https://www .rcsb.org) [51].The structure was selected because it exhibited a lower resolution (<1.90 Å) in its X-ray crystallographic structure and demonstrated better percentile scores in global validation metrics, signifying superior structural quality.
Water molecules, native ligands, and heteroatoms present in the crystallized protein structure were removed using BIOVIA Discovery Studio 2019 software [52].Additional critical factors, such as side-chain geometry, hydrogen correction, and correction of improper bond orders, were addressed and minimized using the GROMOS 43B1 force field of Swiss-PDB Viewer version 4.10 [53].A total of 692 compounds were included as ligands from the NIH clinical collection; these compounds consisted of molecules with a documented history of use in human clinical trials.The 3D structures of these compounds were retrieved from the PubChem database (https://pubchem .ncbi.nlm.nih.gov/)[54].The ligands underwent energy minimization using the mmff94 force field within the Open Babel plug-in of the PyRx tool [55].
2.10.Molecular Docking.Molecular docking was performed by employing the AutoDock plug-in within PyRx software by loading and converting each ligand and the GMFG receptor from the PDB format to the PDBQT format.Hydrogens with polar characteristics were introduced into the enzyme, and hydrogens with nonpolar characteristics were combined.The grid box for the GMFG enzyme was established as follows: the center points of the box were X = 4 8407, Y = 27 8057, and Z = 12 8898, and the dimensions (Å) were X = 46 6266, Y = 42 0797, and Z = 38 1385.Molecular docking was carried out using AutoDock Vina in PyRx [56].The binding affinities of the ligands to the receptors were calculated in kcal/mol, with negative values indicating stronger binding.Molecular visualization and nonbonding interactions of the protein-ligand complexes were analyzed utilizing Discovery Studio 2019.

Study of ADMET and Drug
Likeness.The ADMET and pharmacokinetic properties of the top ligands were predicted by inputting their canonical SMILES sequences into the admetSAR 2 (http://lmmd.ecust.edu.cn/admetsar2)[57] and SwissADME (http://www.swissadme.ch/index.php)servers [58].The canonical SMILES for each lead molecule was obtained from the PubChem database.ADMET properties were determined for each ligand by utilizing various parameters, such as Lipinski's rule of five, BBB permeability, aqueous solubility, toxicity, and carcinogenicity.These characteristics significantly enhance the drug potential and effectiveness of phytochemicals in the treatment of diseases.
2.12.Molecular Dynamics (MD) Simulation.MD simulation is a widely used technique for exploring and validating the structural flexibility of protein-ligand complexes within a present time frame in a controlled environment.It can be applied effectively to reveal dynamic interactions and understand macromolecular structure-to-function relationships [59].YASARA dynamics software was used to confirm the prediction results from the docking study [60].Equilibration and production steps are included in the MD simulation, which starts after complete energy minimization [61].To evaluate the structural integrity of the complexes, the GMFG protein was utilized as a control.The AMBER14 force field was applied in this simulation process, which is widely accepted [62].The PME (particle-mesh Ewald) method was used to compute the long-range electrostatic interactions in the study, and the cut-off radius was fixed at 8 Å [63].In this MD simulation, a 1.25 fs time step was applied, and the trajectories were saved after every 100 ps.The runtime of the MD simulation was 100 ns while keeping the pressure constant and employing the Berendsen thermostat [64].By applying the TIP3P solvation model, a cubic simulation cell was generated, and periodic boundary conditions were maintained [65].The overall environmental conditions for the system were established at a temperature of 298 K, pH 7.4, and a NaCl concentration of 0.9% [66].The system was minimized by utilizing the steepest descent method [67].The root-mean-square deviation (RMSD), root-mean-square fluctuation (RMSF), radius of gyration (RG), hydrogen bond number, and solvent-accessible surface area (SASA) were evaluated by analyzing the trajectory data from MD simulations [68].

Results
3.1.Data Processing and DEG Identification.We analyzed the gene expression data of GBM to explore the DEGs.In comparison to those in normal patients, in the GSE50161 dataset, we found 5405 significant DEGs, 2902 of which were upregulated and 2503 of which were downregulated.This investigation also identified 1796 (827 upregulated and 969 downregulated) and 1479 (760 upregulated and 719 downregulated) DEGs in the GSE12657 and GSE15824 datasets, respectively.The upregulated genes and downregulated genes are visualized in Figures 1(a), 1(b), and 1(c).A comparative assessment of the three datasets revealed 55 common upregulated (Figure 2(a)) and 23 common downregulated DEGs (Figure 2(b)).These 78 common DEGs were subsequently subjected to PPI analysis.
3.2.Hierarchical Clustering Analysis.We selected 78 genes for clustering analysis of gene expression data, chosen for their common presence across our targeted datasets.Figure 3 presents a dendrogram depicting the hierarchical clustering of these selected genes.The color gradient illustrates the degree of gene expression regulation, with three primary color presets for effective visualization.Specifically, red indicates maximum or upregulated values, green represents minimum or downregulated values, and black signifies neutral or zero values, indicating no difference in expression.

PPIs and Identification of Hub DEGs.
We explored PPI networks predicted by the common DEGs related to this disease to understand their functional interactions.The PPI pair contains 76 nodes and 29 edges, and the enrichment p value for the PPI is less than 1 0e − 16.Only interconnected nodes and edges are depicted in Figure 4(a).Moreover, several related nodes in the PPI networks were identified as hub genes.A topological analysis of the networks revealed 9 hub proteins for the common DEGs by using the MCODE algorithm, as shown in Figure 4(b).A list of the hub DEGs is presented in Table 2.

Analysis of Gene Ontologies (GO) and Pathway
Enrichment.After determining the hub DEGs associated with the disease, we conducted significant GO and pathway enrichment analyses using curated databases to investigate the gene ontologies and pathways.Most of the DEGs were involved in the innate immune system.A wide range of GO terms and signaling pathways were enriched.The findings of the pathway analysis showed that the genes were mostly linked to several immune system processes.The top 10 biological processes, molecular functions, cellular components, and pathways are enumerated in Tables 3 and 4.
3.5.Regulatory Signatures Revealing Significant TFs.We identified 10 TFs associated with the targeted DEGs to uncover regulatory biomolecules that are likely to govern the expression of dysregulated genes at the posttranscriptional level (Table 5).Significant TFs, including Pax6, PPARG, INSM1, RREB1, CTCF, REST, NHLH1, RXR::RAR, DR5, RXRA::VDR, and PLAG1, were found to play significant roles in the regulation of the DEGs identified in this study.

Survival Analysis.
The correlation between the level of GMFG expression and overall survival (OS) in GBM patients was analyzed via the Kaplan-Meier plots.According to the p value, there were no correlations between the gene expression of eight genes and OS, with the exception of GMFG.These findings suggest that high expression of GMFG is significantly associated with poor OS (Figure 5).

3.7.
Validation of the GMFG Gene.The GMFG gene was studied to determine the extent to which it was expressed at the mRNA level.Analysis via the GEPIA 2.0 server revealed higher expression of GMFG mRNA in GBM tissues (normal: 207; tumor: 163) than in normal tissues (Figure 6(a)).The GMFG gene was subsequently evaluated on the OncoDB web server to determine its expression patterns in malignant tissues.Like in previous results, GMFG mRNA was more highly expressed in GBM tissues (p = 1 9e − 35) than in normal samples (Figure 6  9 BioMed Research International bond with TYR84, LEU117, and LYS119.Conversely, the GMFG-5′-guanidinonaltrindole interaction formed conventional hydrogen bonds with ARG22, ARG24, and GLU26; pication interactions with ARG22; pi-alkyl interactions with LEU86, LEU117, and PHE121; and alkyl interactions with LYS119 and LEU138 (Figure 7).A surface view and receptor-ligand interactions of the docked complexes for the top two compounds within the active and catalytic sites of the GMFG protein are shown in Figure 8.The binding scores and types of interactions with the target protein can be found in Table 6.
3.9.ADMET Analysis.The various pharmacokinetic parameters of the top 10 identified compounds were evaluated based on Lipinski's Rule of Five.Risperidone and 5 ′ -guanidinonaltrindole were selected because they showed no violation of Lipinski's Rule of Five.The ADMET properties of the selected compounds indicated ideal drug-like characteristics (Table 7).Considering factors such as the lowest binding affinity and pharmacokinetic properties, we selected the top two ligands, risperidone and 5 ′ -guanidinonaltrindole, for further study.
3.10.MD Simulation.MD simulation is an essential analysis of the structural stability and flexibility of biological macromolecules [69].MD simulations were performed for 100 ns to investigate the conformational changes and binding mechanism of the (GMFG) protein, (GMFG)-risperidone complex, and (GMFG)-5 ′ -guanidinonaltrindole complex.After 100 ns of simulation, dynamic trajectories were inspected, and parameters such as the RMSF, RMSD, and RG (Figure 9), as well as the number of hydrogen bonds and solvent accessible surface area (SASA) (Figure 10), were computed.
By analyzing the RMSF of the (GMFG) protein, (GMFG)-risperidone complex, and (GMFG)-5′-guanidinonaltrindole complex, it was observed that the protein and the two complexes displayed an overall similar range of fluctuations.These findings suggested that the two complexes exhibited desirable protein residue flexibility; here, the fluctuation range of the protein was considered a control.To check the firmness of the GMFG protein, the GMFG-risper complex, and the GMFG-5 ′ -guanidinonaltrindole complex, the RMSD values of the Cα atoms of these compounds were assessed [70].The GMFG protein slightly fluctuated from 75 to 81 ns.However, all the other trajectories of the protein remained stable.However, from the start to the end, all the trajectories of both the (GMFG)-risperidone complex and the (GMFG)-5 ′ -guanidinonaltrindole complex exhibited an equilibrium state, which suggested the strong stability of both protein-ligand complexes with respect to the reference protein (GMFG).
The structural compactness and Rg were evaluated [71].The GMFG-risperidone complex was stable at the starting position.However, from 8 to 38 ns, several small fluctuations were observed.After that, the equilibrium state was reached.The (GMFG)-5′-guanidinonaltrindole complex also showed initial stability.However, from 7 to 70 ns, several fluctuations occurred.Afterwards, the material also exhibited stability.Based on the (GMFG) protein Rg trajectory pattern as a control, it was observed that the (GMFG)-risperidone complex had a better Rg profile than the (GMFG)-5 ′ -guanidinonaltrindole complex.The GMFG protein, (GMFG)-risperidone complex, and (GMFG)-5 ′ -guanidinonaltrindole complex had similar numbers of hydrogen bonds, which suggests the desirable integrity of these compounds.SASA was analyzed to evaluate folding and stability [72].The (GMFG)-risperidone complex exhibited a fluctuating nature from 5 to 11 ns.Afterwards, the system reached a more or less stable equilibrium state, which indicated the considerable compactness of the compound.In contrast, the (GMFG)-5 ′ -guanidinonaltrindole complex exhibited high fluctuations from 2 to 10 ns and 46 to 58 ns, which indicated that it had slightly decreased stability.Based on the GMFG protein SASA trajectories as a reference, the (GMFG)-risperidone complex was shown to have a better SASA profile or compactness than the (GMFG)-5 ′ -guanidinonaltrindole complex.
Analysis of these parameters indicated that despite the two complexes showing potentiality and very close outcomes, the (GMFG)-risperidone complex is considered a more promising candidate due to its better Rg and SASA      To demonstrate the changes in the binding cavity, the superimposition of pre-and post-MD simulation structures of risperidone and 5 ′ -guanidinonaltrindole was assessed (Figure 11).The structural changes in risperidone and 5 ′ -guanidinonaltrindole were observed after every 25 ns.A snapshot was taken of the surface view of the (GMFG)-risperidone complex and the (GMFG)-5 ′ -guanidinonaltrindole complex at 25, 50, 75, and 100 ns (Figures 12 and 13).

Discussion
GBM is an aggressive and highly malignant brain tumor with limited treatment options and a poor prognosis.Over the years, significant efforts have been made to better understand the molecular mechanisms underlying GBM development and progression, leading to the exploration of novel treatment strategies.One area of research involves the exploration of molecular and genetic alterations in GBM, intending to identify specific biomarkers and therapeutic vulnerabilities.For instance, studies have revealed the role of genetic mutations in genes such as IDH1, EGFR, TP53, and PTEN in GBM development and progression [1].Additionally, advancements in genomic sequencing techniques have enabled comprehensive profiling of GBM tumors, aiding in the identification of potentially targetable mutations [15].Furthermore, immunotherapy approaches, including immune checkpoint inhibitors and personalized vaccines, are being investigated as promising strategies to harness the immune system's response against GBM [73].Despite numerous research efforts, the precise mechanisms driving the development of GBM have not been fully elucidated.Consequently, there is a pressing need for further research to uncover potential biomarkers and develop more efficient treatment options for GBM, aiming to improve the prognosis and OS rates of individuals affected by this highly destructive condition.
In the present study, we analyzed the gene expression data of GBM patients to identify DEGs, hub DEGs, genes related to biological activities, molecular pathways, regulatory biomolecules, and potential biomarkers.This was performed utilizing a multiomics data integration framework to identify potential therapeutic targets for GBM.By analyzing the patterns of gene expression, we found 8680 DEGs, 4697 of which were upregulated and 3983 of which were downregulated.Comparative analysis of the datasets identified 55 upregulated and 23 downregulated DEGs.Moreover, PPI analysis revealed nine hub DEGs, which encode significant hub proteins that exhibit strong interconnections.These hub DEGs were found to be closely associated with various biological activities and pathways in the functional enrichment study.These activities and pathways included  ) for the hub DEGs.These TFs are believed to play a significant role in the regulation of gene expression in GBM and are considered to be major regulators of this disease.Survival analysis indicated that high expression of the GMFG gene was significantly associated with shorter OS than was high expression of other hub DEGs.Therefore, overexpression of the GMFG gene may lead to unfavorable outcomes in GBM patients.Consequently, we validated these DEGs as potential prognostic biomarkers.
Initially, at the mRNA level, the GMFG gene was shown to be differentially expressed (upregulated) in GBM tissues, indicating its potential tumourigenic role in the formation and development of GBM.Further, we conducted a comparative study on immunohistochemistry between GBM and normal cells.In this investigation, we observed significant differences in staining, intensity, and quantity between tumor cells and normal glial cells.In a previous study, it was observed that GMFG exhibited higher expression levels in GBM tissues [74].This consistency across studies enhances the credibility of the results and strengthens the argument for the role of GMFG in GBM.Since cancer formation is a complex process that varies depending on the subtype, grade, and demographics of the cancer [75], it is necessary to analyze the expression level of a gene across different variables to understand its function in disease development.Then, we evaluated the efficacy of the GMFG protein as a drug target using molecular docking and dynamics studies.
Molecular docking is a valuable technique that involves the binding of a small molecule (ligand) to the binding site of a target receptor to demonstrate its attachment.In this study, a total of 692 ligand molecules were selected to predict 17 BioMed Research International inhibitors of the GMFG protein that are associated with the development of GBM.The ligands were docked against the target receptor to evaluate the anti-GMFG potential of the ligands, and therefore, the two ligands with the best docking were chosen for further analyses.The inhibitory efficacy of the ligands on the receptor was determined by the lower binding affinity score.This study demonstrated that risperidone and 5′-guanidinonaltrindole have strong affinities for GMFG, as indicated by binding scores of −8.5 and −8.4 kcal/mol, respectively.Risperidone is an antipsychotic medicine that is commonly used to treat a variety of psychiatric conditions, including schizophrenia [76], bipolar disorder [77], Alzheimer's disease [78], and dementia [79].Though the risperidone is an antipsychotic drug, it has been shown to have a therapeutic effect on GBM [80].This compound was also reported to have a potential effect for lung 18 BioMed Research International cancer [81] and gastric cancer [82].This drug molecule forms multiple nonbonded interactions with the GMFG protein and interacts with the TYR84, PRO85, LEU86, LEU117, and LYS119 residues.5 ′ -Guanidinonaltrindole is an opioid antagonist that has been shown in animal experiments to have antidepressant effects [83,84].Specifically, it is a selective kappa-opioid receptor (KOR) antagonist that has shown promise in cancer treatment through its ability to modulate various signaling pathways associated with tumor growth and metastasis [85].5′-Guanidinonaltrindole interacted with the ARG22, ARG24, GLU26, LEU86, LEU117, LYS119, PHE121, and LEU138 residues of the GMFG protein.
The results of the Lipinski filter, SwissADME, and admetsar2 analyses demonstrated that the selected top ligand molecules also overcame the drug-like requirements.These evaluations highlighted the possible pharmacokinetic properties of these compounds.
Estimation of pharmacokinetic properties facilitates drug development.The permeability of a drug through biological barriers is influenced by factors such as the topological polar surface area (TPSA) and molecular weight.A larger molecular weight and TPSA are linked to lower drug permeability.BBB penetration is necessary for the use of medications to treat the brain.The selected ligand molecules have the ideal molecular weight, TPSA, optimum number of H-bond donors and acceptors, and ability to penetrate the BBB.Additionally, the toxicity profiles of the two compounds showed that the drug molecules have no carcinogenicity or toxicity, suggesting their potency as drugs.
MD simulation is a powerful tool for assessing postmolecular docking analysis and validation [86].Risperidone and 5 ′ -guanidinonaltrindole were the selected compounds after docking, and ADMET analysis was performed.Therefore, the (GMFG)-risperidactam (risperidone) complex and the (GMFG)-5 ′ -guanidinonaltrindole complex were evaluated via 100 ns MD simulation, where the simulation of the GMFG protein was used as a reference.The parameters of the RMSF, RMSD, RG, hydrogen bond number, and SASA were analyzed thoroughly.Both complexes were stable following the reference protein.However, the (GMFG)risperidone complex displayed more desirable Rg and SASA results than did the (GMFG)-5′-guanidinonaltrindole complex.Therefore, risperidone is most likely the best inhibitor of the target protein.
The tendency of the GMFG gene to be overexpressed and its effect on survival patterns suggest that the GMFG gene can be used to diagnose and treat GBM.Overall, the findings of this empirical study acquired from differential expression analysis, promoter methylation, and mutation rate of the GMFG gene in GBM tissues indicate that this gene is most likely to play a crucial role in the development and treatment of GBM.Finally, this study suggested that the GMFG is likely to serve as a viable prognostic and therapeutic target for GBM.The clinical relevance of GMFG as a prognostic biomarker and therapeutic target needs to be validated in larger and independent cohorts of GBM patients.The actual biological activity and therapeutic efficacy of the identified compounds need to be confirmed through labora-tory and clinical experiments.Although our molecular docking and dynamics simulation substantiated a preliminary result of the interaction between GMFG and the candidate compounds, these finding needs to be validated by clinical trials with GBM patients.

Conclusion
Analysis of the hub DEGs by GO and pathway enrichment revealed the important processes and pathways involved in GBM.Furthermore, this study demonstrated a significant correlation between the overexpression of the GMFG gene and OS in GBM patients.Our research also revealed that the expression level of GMFG is high in GBM tissues, suggesting that GMFG is a potential diagnostic and therapeutic target for GBM.Risperidone and 5 ′ -guanidinonaltrindole exhibited significant docking energy when interacting with two lead molecules.The MD simulation also validated the stability of the interaction between the active pocket of the protein and the compounds.Overall, our findings revealed that risperidone is a potential target for GBM treatment.This investigation holds promise for advancing the understanding and management of GBM and, thus, can aid in disease treatment and drug development.

Figure 4 :
Figure 4: PPI network for identifying the hub DEGs.(a) Protein-protein interactions of the common DEGs.(b) Network of hub DEGs determined by the MCODE algorithm.

Figure 6 :
Figure 6: The mRNA expression level of the GMFG gene in GBM tissues.(a) GEPIA2 server.The black and red boxes represent normal and cancerous samples, respectively (log2 transformation was used to normalize the results).(b) OncoDB server.(c) UALCAN server.Based on immunohistochemistry data, the protein expression levels of GMFG in (d) normal tissue (glial cells) and (e) GBM (tumor cells) are provided.
profile.As determined by MD simulation, minor conformational changes occurred in these two complexes, similar to what was observed for the GMFG reference protein.

Figure 11 :
Figure 11: The superimposed view of pre and postmolecular dynamics simulation structures of (a) risperidone and (b) 5′-guanidinonaltrindole.The light in green denotes the premolecular dynamic structure, and the hot pink denotes the postmolecular dynamic structure.

Figure 12 :Figure 13 :
Figure 12: Surface view of the risperidone and GMFG protein complex after 25, 50, 75, and 100 ns of MD simulation.The figure shows the structural change every 25 ns.

Table 1 :
Details of the selected datasets, including GEO accession numbers, platforms, case and control samples, and DEG counts.

Table 2 :
The predicted hub DEGs with their designation.

Table 3 :
Gene Ontology analysis of the hub DEGs of glioblastoma.The top 10 terms of each category are listed herewith.

Table 4 :
Pathway analysis of the common DEGs.The top 10 terms of each category are listed.

Table 5 :
Top regulatory molecules (TFs) for common DEGs of glioblastoma.

Table 7 :
Pharmacological profile and drug-likeness studies of the top 10 potential compounds for GMFG targets.