INTRODUCTION
Pulmonary hypertension is generally named essential (idiopathic) or secondary. It is presently clear, in any case, that there are circumstances inside the classification of optional pneumonic hypertension that look like essential pneumonic hypertension in their histopathological highlights and their reaction to treatment (Simonneau et al., 2019). Pulmonary blood vessel hypertension is characterized as a consistent raise of pulmonary arterial strain to more than 25 mm Hg very still or to in excess of 30 mm Hg with workout, with a mean pneumonic slender wedge weight and left ventricular end-diastolic weight of under 15 mm Hg (Frost et al., 2019).
Pneumonic blood vessel hypertension involves idiopathic aspiratory blood vessel hypertension (once in the past, essential aspiratory hypertension), pulmonary blood vessel hypertension in the setting of collagen vascular sickness [e.g., in restricted cutaneous foundational sclerosis, otherwise called the Calcinosis, Raynaud’s, esophageal dysmotility, sclerodactyly, telangiectasia (CREST) condition, entryway hypertension, inborn left-to-right intracardiac shunts, and disease with the human immunodeficiency infection and determines pneumonic hypertension of the infant (Noordegraaf et al., 2019)]. Intimal fibrosis increased medial size, pulmonary arteriolar blockage, and plexiform lesions are all histological features of lung tissue in each of these patients. While the pathophysiology of many pulmonary arterial hypertension (PAH) procedures is unknown, numerous recent discoveries have been made, particularly in the area of genetic and biological cells in idiopathic PAH (Pagnesi et al., 2020).
Pulmonary hypertension (PAH) is a common global disease in which irreversible changes in blood vessels, resulting in long-term resistance of blood vessels and right ventricular failure, are caused. the last, death. Currently, researchers have been able to study the physiological behavior of this disease and its impact on human health (Delcroix et al., 2021). Recently progress has been made toward the understanding of basic pathological biology and symptoms of PAH disease, biomarker prognosis, and treatment options. However, studies are still scarce which provide PAH-related gene expression profiles. Consequently, identifying clinical molecular biomarkers and probing the underlying biology involved in PAH is an important mission that can help develop novel science-based diagnostics and adopt goal-treatment approaches in PAH patients (Vachiéry et al., 2018).
Microarray technology is the most widely used gene expression technology used to study simultaneous expression and regulation of thousands of genes at cellular, organ, or organism level. This novel strategy aids in a more systematic understanding of gene regulation and connections between genes. In microarray practice, however, many undesirable systematic changes are observed (Yadav and Srivastava, 2018). Even in the repeated experiment, some variations are observed. Normalization is the technique of removing specific variances that affect gene expression levels measurement. Despite the fact that several standard ways have been presented, determining which method works best is tough. In the early stages of microarray data analysis, synchronizing is crucial. The results of subsequent analyses are heavily reliant on normalization (Yadav and Srivastava, 2016).
A variety of novel experiments have been developed to study and perform comparison of two samples (e.g., tumor versus normal tissue) based on their gene expression. Microarray is such experiment that measures the intensity values that is proportional to gene expression and it is widely used for comparative study (Yadav and Srivastava, 2019). Certain aspects of microarray investigations enable the development of such novel methods: (i) the enormous number of genes that contribute to expression measurements, which much outweighs the number of samples (observations) available, and (ii) the fact that gene expression measurements are usually highly correlated when it comes to pathway / network interactions. These issues are exacerbated in regression situations, where the goal is to link the expression of several genes to an external outcome or phenotype at the same time. As a result, numerous approaches to addressing these concerns have lately been presented. This shows that gene collecting, without any further constraints, can take up to a year. Using a microarray-based method for evaluating cardiomyopathy in transgenic mice has been widely used for the identification of disease-causing gene and its functional behavior (Basavegowda and Dagnew, 2020).
Expression of genes, a glimpse of all transcriptional activity in a biological sample is provided by microarrays. Microarrays enable the discovery of new and unexpected gene roles, unlike most standard molecular biology methods, which typically allow the analysis of a single gene or a limited number of genes. These technologies have been used to find novel disease subgroups, build new diagnostic tools, and identify basic illness or treatment response mechanisms, among other things. However, because this technology inevitably generates a vast amount of data, we must interpret it using sophisticated statistical and computational tools (Leday et al., 2018).
Microarrays detect levels of expression in relation to measuring mRNA resistance in thousands of genes that are not embedded in a glass (“chip”) gene, providing a unique way of finding the balance and integration produced by genetic sequencing efforts. From inverted northern blots on filters detected with radioactive probes to a high-tech field including minorized synthesis, multicolor fluorescence identification, and database-driven sample and data management, microarray technology has advanced (Mehmood et al., 2017). Whole genomes have recently been analyzed, as well as specific gene families. Gene selection, microarray synthesis, sample preparation, array hybridization, detection, and data processing are typical procedures in microarray tests, with the relevant controls required for each step. Microarray technology is also known as high-bandwidth technology used for the parallel analysis of gene expression for several genes of known and unknown function (Sayed et al., 2019).
MATERIALS AND METHODS
Microarray data of PAH disorders was retrieved from NCBI Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo) with the accession number GSE113439 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE113439). Microarray experiment data used for current research is of PAH diseases. This experiment has been conducted using GPL6244 Affymetrix Human Gene 1.0 ST Array platform and it includes 15 PAH samples from patients and 11 samples from normal controls persons. From 15, only 6 PAH patient’s samples and these samples were labeled as Group 1 were taken for the research work. Out of 11 normal controls, 6 samples were taken as control samples and labeled as Group 2. Differential gene expression analysis was done between Group 1 (PAH patient’s samples) and Group 2 samples (normal controls). The workflow used for the analysis and identification of hub genes has been shown in Figure 1. It shows the list of Bioconductor packages, tools, and databases used for the current study this pipeline can be used for the prediction of hub genes from microarray data and also for network construction.
R and Bioconductor packages (https://www.bioconductor.org/) were used for the statistical analysis of microarray data. Affy and AffyQCReport (Parman et al., 2020) is a package of R functions and analysis classes of oligonucleotide arrays produced by Affymetrix. The normalization of data was done using the Affy Bioconductor package. First the raw data, i.e., CEL files were read using the ReadAffy () command. The data is stored in an affybatch object which is further called in several analysis. Robust Multi-array Average (RMA) method of normalization was applied on the microarray data. RMA is a normalization procedure for microarrays that background corrects, normalizes, and summarizes the probe level information without the use of the information obtained in the MM probes. The expression file is generated after doing the normalization of data.
AffylmGUI (quality assessment) AffylmGUI Package (Harris et al., 2021 affylmGUI is a graphical user interface (GUI) in the integrated workflow of Affymetrix microarray data. This package was used for complete analysis of microarray data from quality control analysis to the application of linear fit model between different set of samples.
 | Figure 1. Workflow for construction of network from microarray gene expression data. Workflow represents the methodology used along with databases and tools used in each step. [Click here to view] |
Limma package Limma is a R / Bioconductor software package that can be used for the identification of differentially expressed genes (DEGs) using linear model (Ritchie et al., 2015). Limma package Limma was performed to identify the top DEG using the linear fit model. For the Limma analysis, different contrast between the samples is made so as to obtain the DEG between them. Top genes file was generated through Limma package and volcano plot was constructed through it. Statistical analysis of data was done using Limma package.
>AffyBatch object
>size of arrays = 1,050 × 1,050 features (23 kb)
>cdf = HuGene-1_0-st-v1 (32,321 affyids)
>number of samples = 12
>number of genes = 32,321
>annotation = hugene10stv1
> library (limma)
> colnames (design) <- c (“group1”, “group2”)
> fit = lmFit (eset, design)
> contrast.matrix <- makeContrasts (group2-group1, levels = design)
> fit21 = contrasts.fit (fit, contrast.matrix)
> fit21 = eBayes (fit21)
> table21 = topTable (fit21, coef=1, adjust=”BH”)
> table21
> top_genes <- topTable (fit2, number = 100, adjust = “BH”)
> top_genes
The network created through string database was visualized using Cytoscape (https://cytoscape.org/) (Otasek et al., 2019) and the protein—protein interaction (PPI) network studied thoroughly. CytoHubba tool (https://apps.cytoscape.org/apps/cytohubba) (Ma et al., 2021) was used for the identification of hub genes. All the network created for both the genes which includes significant genes of T-test analysis and the top 25 genes from Limma Bioconductor package were subjected to cytoHubba tool of Cytoscape.
RESULT AND DISCUSSION
Affy QC plots
The density plot and the Raw boxplot was obtained through affylmgui package (Figs. 2 and 3) these plots show the variation in intensity across chip of all probes. Density plots represent the log intensity of chip expression in in x-axis and y-axis shows the density of chip expression in all sample files. Box plot shows the variation in intensity measured across chip using five different values (minutes, first quartile, median, second quartile and maximum values). Box plot x-axis represent the sample files under consideration for current study compared against one another across five values represented in y-axis. Before statistical analysis normalization was done and normalized intensity Box Plot for each array has been shown in Figure 4. RNA degradation plot was also studied to visualize the intensity of probes from 5? end to 3? end and shown in Figure 5. RNA degradation plot shows the intensity difference of probes those located at the 5? end and 3? end of mRNA. This plot directly measures the quality of RNA probes over microarray chip and shows that intensity of probes degrades towards 3? end. Hence, this parameter is important for differential gene expression analysis and to avoid biasness in interpretation of the results.
Limma package
Differentially expressed genes were identified by R and Bioconductor Limma package with default parameters. Top 25 genes were selected based on linear fit model analysis using Limma package and ranking of genes according to B value. Further functional annotation, these genes were also used for the prediction of PPI and for the identification of hub genes. A volcano plot was visualized that shows the scatterplot between statistical significance which is p-value versus magnitude of change which fold change. Figure 6 shows the Volcano plot between two samples used for the identification of DEGs. Volcano plot analysis and according to p-value it is clear that there is no significant difference in gene expression between PAH samples and control samples. Hence, protein interaction and identification of hub genes becomes crucial in further investigation and analysis.
 | Figure 2. Density plot shows the variation between all sample’s files. [Click here to view] |
 | Figure 3. Raw boxplot shows the variation across all files using five different values (minutes, first quartile, median, second quartile and maximum values). [Click here to view] |
Detailed description of the top 25 DEGs used for further study has been shown in Table 1, it describes the gene name, statistical values like log FC values, t value, p value etc. These 25 genes have been selected on the basis of linear fit model statistics applied using LIMM package and ordered according to B value that is log -odds value of DEGs.
 | Figure 4. Normalized intensity Box Plot for each array through afflmGUI shows all the samples files have been normalized across all five values. [Click here to view] |
 | Figure 5. RNA degradation plot through affylmGUI. It shows the intensity from 5? end to 3? end of probes. [Click here to view] |
 | Figure 6. Volcano plot between two samples for the identification and visualization of DEGs. [Click here to view] |
For prediction of significance and mechanism of these 25 genes, further protein interaction and functional enrichment was done. PPI study of all 25 genes predicted from Limma package has been used to study protein interaction using STRING database (https://string-db.org/) (Szklarczyk et al., 2019). PPI result has been shown in Figure 7 with the following network properties nodes: 26, edges: 20, node degree (average): 1.54, local clustering coefficient: 0.307 and PPI enrichment p-value: 0.0582. Six proteins were identified outside the network and they do not have any edges these are the proteins that do not have any interacting partner.
Further PPI network was visualized in Cytoscape tool for further analysis and prediction of hub genes as shown in Figure 8. CytoHubba tool was used for the identification of hub genes as shown in Figure 9.
Five proteins were identified as hub genes out of the top 25 DEGs which was obtained from Limma package (Fig. 9). Functional annotation and enrichment of all five hub genes was done using Gene Ontology (GO) database (http://geneontology.org), Kyoto Encyclopedia of Genes and Genomes (KEGG) database (https://www.genome.jp/kegg/pathway.html), etc. Hub genes annotation with their Gene Name, GO ID, GO Molecular Function, and KEGG Pathways have been shown in Table 2.
 | Table 1. Top 25 DEGs identified from Limma Package. Table shows the affymetrix id, gene name along with the statistical values like log FC values, t value, etc. [Click here to view] |
Serine/threonine-protein kinase RIO1 (RIOK1)
Function
Involved in the final stages of the 40S ribosomal subunit’s cytoplasmic formation and the conversion of 18S-E pre-rRNA to mature 18S rRNA. Reprocessing of NOB1 and PNO1 from the late 40S precursor is required. Prior to binding to the 60S ribosomal subunit, the connection with the late 40S subunit intermediate may comprise a translation-like checkpoint point cycle. Despite the fact that the protein kinase domain is thought to function primarily as an ATPase, (Lim et al., 2019).
Pinin (PNN)
Function
The E-box 1 core region of the E-cadherin promoter gene is bound by transcriptional activator. 5′CAGGTG-3′ is the core-binding sequence. CTBP1-mediated transcription repression was successfully removed. At the splice junction on mRNAs, an auxiliary constituent of the splicing-dependent multiprotein exon junction complex (EJC), is located. The EJC is a dynamic assembly of core proteins and various outlying nuclear and cytoplasmic associated components that only join the compound for a brief period of time, either during EJC construction or after mRNA metabolism. Involved in epithelia cell-cell attachment creation and maintenance. Renal cell carcinoma has a latent tumor suppressor (Meng et al., 2021).
Nucleolin (NCL)
Function
Growing eukaryotic cells have a significant nucleolar protein. It is shown to be linked to pre-ribosomal particles and intra-nucleolar chromatin. By binding histone H1, it promotes chromatin decondensation. It is assumed to be involved in the writing of pre-rRNA, ribosome assembly, and transcriptional writing development. It binds to RNA oligonucleotides with the configuration 5′-UUAGGG-3, which is twice as strong as telomeric single-stranded DNA with the sequence 5′-TTAGGG-3′ (Wang et al., 2019a).
 | Figure 7. Protein-protein network construction of 25 DEGs using STRING database. [Click here to view] |
Eukaryotic translation initiation factor 5B (EIF5B)
Function
The translation process starts. Translational GTPase combines 40S and 60S subunits to form an 80S initiation complex with initiator methionine-tRNA in a P-site base bound to the original codon. This results in consistent changes to an enzyme due to GTP binding and -hydrolysis, resulting in the production of productive interactions with the ribosome. The release of the enzyme is required for the production of strong ribosomes after the initial structure has been formed (Wang et al., 2019b).
Ribosomal L1 domain-containing protein 1 (RSL1D1)
Function
The regulation of cellular ageing is done through inhibition of translation of PTEN. In response to DNA damage, it acts as a pro-apoptotic regulator. EIF5B is highly divergent and belongs to the common ribosomal protein uL1 family (Cheng et al., 2020).
DEGs identified through DEGs analysis of PAH and control samples were used for the construction of PPI map. PPI map and with was used for the prediction of functional and hub genes that are expressed in PAH disease condition. Five genes that are EIF5B, NCL, PNN, RIOK1, RSL1D1 were identified as hub genes and have functions related with PAH disease. Functional analysis of these five hub genes shows that they have function in cancer (RSL1D1), RNA transport (EIF5B, PNN) etc. Research has been done to study the mechanism and genes expressed in PAH condition, it has been shown that genes are expressed in early stage of embryo development and hence some fetal genes are also responsible (Lemay et al., 2021).
Genes associated with PAH condition has also been shown in connective tissue diseases and many genes has been identified like NOCH 1, TBX4 etc (Hernandez-Gonzalez et al., 2021). Microarray gene expression experiments has also been done to identify drug targets like CCR1, CCL5, SAA1 etc. (Xu et al., 2021). These findings along with the current study and results provides more insight into the PAH diseases that can be used for molecular diagnostics and in pharma industry for the designing of drugs. Identified genes can be used as novel drug targets and towards the molecular understanding of PAH.
 | Figure 8. PPI visualized in Cytoscape tool and identification of hub genes highlighted in red circle with yellow nodes. [Click here to view] |
 | Figure 9. Top 5 hub genes identified using CytoHubba tool. [Click here to view] |
 | Table 2. GO, molecular function, and pathways of all five hub genes. [Click here to view] |
CONCLUSION
Microarray gene expression experiment is an important tool to study the differential gene expression in different state or functional category of cell. This experiment can be used in multidimensional analysis for the identification of DEGs, co-expressed genes, pathway analysis, protein interaction analysis etc. In this research we have used microarray data for the identification of hub genes that are expressed in PAH patients. The microarray data analysis was done on the data of PAH patients and their control retrieved from public database of microarray. DEGs were identified using R and Bioconductor packages like Limma, Affy package etc. Statistical test like T-test analysis was done using Limma package and top 25 DEGs were further used for the study. These 25 DEGs were used for the network construction through STRING database. Then the network was redirected to Cytoscape where the study of PPI was done and then the five hub genes were identified through cytoHubba. Hub genes identified out of 25 genes were: EIF5B, NCL, PNN, RIOK1, RSL1D1. Function annotation of these five hub genes shows that they have function in cancer (RSL1D1), RNA transport (EIF5B, PNN), etc. These genes can be useful for the identification of novel drug targets and function understanding of PAH.
ACKNOWLEDGMENT
The authors would like to acknowledge Amity Institute of biotechnology, Amity University Uttar Pradesh, Lucknow campus for providing us facilities to conducting this study. This research project is not funded by any specific grant from funding agencies in the public, commercial, or non-profit sectors.
CONFLICT OF INTEREST
Authors do not have any conflict of interests.
FUNDING
No funds were provided for this research.
AUTHOR CONTRIBUTIONS
All authors made substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data; took part in drafting the article or revising it critically for important intellectual content; agreed to submit to the current journal; gave final approval of the version to be published; and agree to be accountable for all aspects of the work. All the authors are eligible to be an author as per the international committee of medical journal editors (ICMJE) requirements/guidelines.
ETHICAL APPROVALS
This study does not involve experiments on animals or human subjects.
DATA AVAILABILITY
All data generated and analyzed are included within this research article.
PUBLISHER’S NOTE
This journal remains neutral with regard to jurisdictional claims in published institutional affiliation.
REFERENCES
Basavegowda HS, Dagnew G. Deep learning approach for microarray cancer data classification. CAAI Trans Intell Technol, 2020; 5:22–33. CrossRef
Cheng R, Qi L, Kong X, Wang Z, Fang Y, Wang J. Identification of the significant genes regulated by estrogen receptor in estrogen receptor-positive breast cancer and their expression pattern changes when tamoxifen or fulvestrant resistance occurs. Front Gene, 2020; 11.pp: 1117 CrossRef
Delcroix M, Torbicki A, Gopalan D, Sitbon O, Klok FA, Lang I, Jenkins D, Kim NH, Humbert M, Jais X, Noordegraaf AV. ERS statement on chronic thromboembolic pulmonary hypertension. Eur Respir J, 2021; 57: 1–37. CrossRef
Frost A, Badesch D, Gibbs JS, Gopalan D, Khanna D, Manes A, Oudiz R, Satoh T, Torres F, Torbicki A. Diagnosis of pulmonary hypertension. Eur Respir J, 2019; 53: 1–12 CrossRef
Harris GM, Abbas S, Miles MF. GCSscore: an R package for differential gene expression analysis in Affymetrix/Thermo-Fisher whole transcriptome microarrays. BMC Genom, 2021; 22:1–2. CrossRef
Hernandez-Gonzalez I, Tenorio-Castano J, Ochoa-Parra N, Gallego N, Pérez-Olivares C, Lago-Docampo M, Palomino Doza J, Valverde D, Lapunzina P, Escribano-Subias P. Novel genetic and molecular pathways in pulmonary arterial hypertension associated with connective tissue disease. Cells, 2021; 10:1488. CrossRef
Leday GG, Vértes PE, Richardson S, Greene JR, Regan T, Khan S, Henderson R, Freeman TC, Pariante CM, Harrison NA, Bullmore E. Replicable and coupled changes in innate and adaptive immune gene expression in two case-control studies of blood microarrays in major depressive disorder. Biol Psych, 2018; 83:70–80. CrossRef
Lemay SE, Awada C, Shimauchi T, Wu WH, Bonnet S, Provencher S, Boucherat O. Fetal gene reactivation in pulmonary arterial hypertension: good, bad, or both? Cells, 2021; 10:1473. CrossRef
Lim H, He D, Qiu Y, Krawczuk P, Sun X, Xie L. Rational discovery of dual-indication multi-target PDE/Kinase inhibitor for precision anti-cancer therapy using structural systems pharmacology. PLoS Comput Biol, 2019; 15:e1006619. CrossRef
Ma H, He Z, Chen J, Zhang X, Song P. Identifying of biomarkers associated with gastric cancer based on 11 topological analysis methods of CytoHubba. Sci Rep, 2021; 11:1. CrossRef
Mehmood R, El-Ashram S, Bie R, Dawood H, Kos A. Clustering by fast search and merge of local density peaks for gene expression microarray data. Sci Rep, 2017; 7:1–7. CrossRef
Meng XY, Zhang HZ, Ren YY, Wang KJ, Chen JF, Su R, Jiang JH, Wang P, Ma Q. Pinin promotes tumor progression via activating CREB through PI3K/AKT and ERK/MAPK pathway in prostate cancer. Am J Cancer Res, 2021; 11:1286.
Noordegraaf AV, Chin KM, Haddad F, Hassoun PM, Hemnes AR, Hopkins SR, Kawut SM, Langleben D, Lumens J, Naeije R. Pathophysiology of the right ventricle and of the pulmonary circulation in pulmonary hypertension: an update. Eur Respir J, 2019; 53: 1–13 CrossRef
Otasek D, Morris JH, Bouças J, Pico AR, Demchak B. Cytoscape automation: empowering workflow-based network analysis. Gen Biol, 2019; 20:1–5. CrossRef
Pagnesi M, Baldetti L, Beneduce A, Calvo F, Gramegna M, Pazzanese V, Ingallina G, Napolano A, Finazzi R, Ruggeri A, Ajello S. Pulmonary hypertension and right ventricular involvement in hospitalised patients with COVID-19. Heart, 2020; 106:1324–31. CrossRef
Parman C, Halling C, Gentleman R. AffyQCReport: QC report generation for affyBatch objects. R package version 1.68.0, 2020. It is bioconductor package and citation has been written as mentioned by developer https://129.217.206.11/packages/3.12/bioc/html/affyQCReport.html
Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucl Acids Res, 2015; 43:e47. CrossRef
Sayed S, Nassef M, Badr A, Farag I. A nested genetic algorithm for feature selection in high-dimensional cancer microarray datasets. Exp Syst Appl, 2019; 1:233–43. CrossRef
Simonneau G, Montani D, Celermajer DS, Denton CP, Gatzoulis MA, Krowka M, Williams PG, Souza R. Haemodynamic definitions and updated clinical classification of pulmonary hypertension. Eur Respir J, 2019;53: 1–13. CrossRef
Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, Simonovic M, Doncheva NT, Morris JH, Bork P, Jensen LJ. STRING v11: protein–protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucl Acids Res, 2019; 47:D607–13. CrossRef
Vachiéry JL, Delcroix M, Al-Hiti H, Efficace M, Hutyra M, Lack G, Papadakis K, Rubin LJ. Macitentan in pulmonary hypertension due to left ventricular dysfunction. Eur Resp J, 2018; 51. CrossRef
Wang F, Zhou S, Qi D, Xiang SH, Wong ET, Wang X, Fonkem E, Hsieh TC, Yang J, Kirmani B, Shabb JB. Nucleolin is a functional binding protein for salinomycin in neuroblastoma stem cells. J Am Chem Soc, 2019a; 141:3613–22: 1–9. CrossRef
Wang J, Johnson AG, Lapointe CP, Choi J, Prabhakar A, Chen DH, Petrov AN, Puglisi JD. eIF5B gates the transition from translation initiation to elongation. Nature, 2019b; 573:605–8. CrossRef
Xu J, Yang Y, Yang Y, Xiong C. Identification of potential risk genes and the immune landscape of idiopathic pulmonary arterial hypertension via microarray gene expression dataset reanalysis. Genes, 2021; 12:125. CrossRef
Yadav R, Srivastava P. Clustering, pathway enrichment, and protein-protein interaction analysis of gene expression in neurodevelopmental disorders. Adv Pharmacol Sci, 2018; 2018: p.3632159 CrossRef
Yadav R, Srivastava P. Significant analysis of microarray (SAM) to identify synergistic effect of RV and NGF in repairing damaged neuronal cells. Toxicol Int, 2019; 25:26–39.
Yadav R, Srivastava P. Visualization of high throughput genomic data using R and Bioconductor. J Data Mining Genom Proteom, 2016; 7:2153–0602. CrossRef