Drug repurposing for coronavirus (SARS-CoV-2) based on gene co-expression network analysis

Severe acute respiratory syndrome (SARS) is a highly contagious viral respiratory illness. This illness is spurred on by a coronavirus known as SARS-associated coronavirus (SARS-CoV). SARS was first detected in Asia in late February 2003. The genome of this virus is very similar to the SARS-CoV-2. Therefore, the study of SARS-CoV disease and the identification of effective drugs to treat this disease can be new clues for the treatment of SARS-Cov-2. This study aimed to discover novel potential drugs for SARS-CoV disease in order to treating SARS-Cov-2 disease based on a novel systems biology approach. To this end, gene co-expression network analysis was applied. First, the gene co-expression network was reconstructed for 1441 genes, and then two gene modules were discovered as significant modules. Next, a list of miRNAs and transcription factors that target gene co-expression modules' genes were gathered from the valid databases, and two sub-networks formed of transcription factors and miRNAs were established. Afterward, the list of the drugs targeting obtained sub-networks' genes was retrieved from the DGIDb database, and two drug-gene and drug-TF interaction networks were reconstructed. Finally, after conducting different network analyses, we proposed five drugs, including FLUOROURACIL, CISPLATIN, SIROLIMUS, CYCLOPHOSPHAMIDE, and METHYLDOPA, as candidate drugs for SARS-CoV-2 coronavirus treatment. Moreover, ten miRNAs including miR-193b, miR-192, miR-215, miR-34a, miR-16, miR-16, miR-92a, miR-30a, miR-7, and miR-26b were found to be significant miRNAs in treating SARS-CoV-2 coronavirus.

. The overall workflow of the proposed method. In this method, a network-based approach is applied to drug repurposing for coronavirus disease treatment. (a) At first, a transcriptome profile for healthy (control) and SARS-CoV-infected samples were taken from the GEO database with the accession number GSE1739. (b) Then, after identifying differentially expressed genes in the control and disease groups, the gene co-expression network is reconstructed, and two significant gene modules are discovered from the co-expression network. (c) Next, for every gene module, the TF-miRNA-TG network is reconstructed independently. The information of TFs-miRNAs, TFs-TGs, and miRNAs-TGs regulations are taken from the TransmiR 42 , TRRUST 39 , and miRWalk 40  www.nature.com/scientificreports/ After analyzing the network in Cytoscape software 43 , generally, 391 nodes and 1273 edges were observed in the reconstructed co-expression network. After clustering the gene co-expression network, we discovered two significant modules (Module A and Module B) in the co-expression network. The list of genes for these two modules is reported in Table 1.

Figure 2.
Gene co-expression network for the primary genes (disconnected genes were removed from the network). The size of the nodes indicates its degree. There are 391 nodes and 1273 edges in this network. DHX15 is the highest degree node in this network. Table 1. Gene names of modules A and B. Numbers inside the parentheses represent the genes degree.
Enrichment analysis of genes. Gene ontology was performed for the module A and B gene lists, separately. The results for module A gene list show that they significantly enriched in mitotic cell cycle process, nuclear division, and mitotic nuclear division biological processes. Moreover, the gene list of module B significantly enriched in RNA processing, mRNA metabolic process, and RNA metabolic process biological processes. Then, a pathway enrichment analysis was done for modules A and B gene lists separately. The results showed that the module A and B genes, significantly enriched in Resolution of Sister Chromatid Cohesion and mRNA Splicing-Major Pathway pathways, respectively. More details of the GO and pathway enrichment analyses are reported in supplementary file S3. Enrichment analysis of miRNAs. In order to check miRNA family for the TF-miRNA-TG_A and TF-miRNA-TG_B sub-networks, at first the list of miRNAs are imported into the TAM online tool. Then, the obtained result is reported in supplementary file S4. As reported in this file, both sub-networks significantly enriched in the let-7 and mir-17 families.

Drug-Gene interaction network.
After gathering drug-gene interactions for TF-miRNA-TG_A and TF-miRNA-TG_B genes, the drug-gene interaction network was reconstructed and is demonstrated in Fig. 3. As shown in this network, some drugs have a high degree, which means that these drugs target and regulate more www.nature.com/scientificreports/ genes. Therefore, high-degree drugs were selected and reported as significant, as they regulate more genes in module_A and module_B. In this regard, the drugs with a degree of 3 or higher were selected and are reported in Table 4. As reported, FLUOROURACIL, EPIRUBICIN, and FOSTAMATINIB are effective drugs and targeted 4, 3, and 3 genes, respectively. www.nature.com/scientificreports/ In addition to the Drug-gene interaction network, the drug-TF network was reconstructed to discover the drugs that target the transcription factors of the TF-miRNA-TG_A and TF-miRNA-TG_B sub-networks. This network is shown in Fig. 4. In this network, the drugs with a degree of 4 and higher were selected and reported as significant drugs. More details of the effective drugs are reported in Table 5. As the table shows, CISPLATIN, SIROLIMUS, and CYCLOPHOSPHAMIDE drugs targeted 11, 6, and 5 TFS, respectively, and the others targeted 4 TFs. The complete information of Drug-Gene and Drug-TF interactions are provided in supplementary file S5. Additionally, supplemental file S6 has detailed information on the network topological characteristics of the Drug-Gene and Drug-TF networks.
Recently, the potential effect of 6710 drugs as SARS-CoV-2 inhibitors is tested in vitro and in vivo 44 . The results of this report show that some of the drugs proposed in this article (see Tables 4 and 5) have an inhibitory effect on SARS-CoV-2. According to this report, OLAPARIB, NICLOSAMID and METHOTREXATE have a very weak, weak, and strong effects on SARS-CoV-2, respectively. As well as, DAUNORUBICIN and BORTEZOMIB have a cytotoxic effect on this disease.

GSEA and candidate drugs validation.
To evaluate the proposed drugs, the Connectivity Map(CMAP) analysis was utilized. To do this, from the Drug-Gene and Drug-TF networks, the drugs with a degree of 4 or above were selected and imported to Enrichr CMAP database. Then, the impact of these drugs on target genes was evaluated. Enrichr CMAP database contains CMAP-up and CMAP-down datasets. In Tables 3 and 4, from among the 19 repurposed drugs for SARS-CoV disease, only nine drugs, including SIROLIMUS, METHYL-DOPA, VORINOSTAT , PACLITAXEL, DAUNORUBICIN, METHOTREXATE, NICLOSAMIDE, and ETOPO-SIDE, were validated by CMAP analysis. Table 6 shows the validated drugs together with the corresponding unregulated or downregulated target genes.  www.nature.com/scientificreports/

Discussion
In this article, a network-based approach was applied to discover therapeutic drugs for SARS-CoV-2 disease. Due to the high genetic similarity of SARS-CoV and SARS-CoV-2, the results of the study on SARS-CoV can be a clue to the treatment of SARS-CoV-2. To this end, at first, differentially expressed significant genes (P-value < 0.05) in healthy and SARS-CoV infected samples were selected, then the gene co-expression network was reconstructed in STRING database for the filtered genes. After reconstructing the gene co-expression network, we discovered two significant gene modules using ClusterViz plugin in Cytoscape software. These two obtained gene co-expression modules (Module A and Module B) contained 25 and 30 genes, respectively. In the next step, a list of TFs and miRNAs which regulate these module's genes were gathered from the TRRUST and miRWalk2.0 databases, respectively. Moreover, the regulation information of the TFs and miRNAs were obtained from the TransmiR database. After collecting the TFs, miRNAs, and TGs regulation information, the two sub-networks of TF-miRNA-TG_A and TF-miRNA-TG_B were reconstructed. TF-miRNA-TG_A contains 347 miRNAs, 25 TGs, and 57 TFs, and TF-miRNA-TG_B contains 116 miRNAs, 25 TGs, and 4 TFs.
To analyze the biological processes and pathways in which Module A and Module B are involved, Gene Ontology (GO) and pathway enrichment analyses were done using a DAVID tool. The GO enrichment analysis showed that the genes in Module A significantly enriched in mitotic cell cycle process, nuclear division, and mitotic nuclear division biological processes. In Module B, the genes significantly enriched in RNA processing, mRNA metabolic process, and RNA metabolic process terms. The Reactome pathway enrichment analysis showed that modules A and B enriched in Resolution of Sister Chromatid Cohesion and mRNA Splicing-Major Pathway terms, respectively (see supplementary file S3). To analyze the TF-miRNA-TG_A and TF-miRNA-TG_B sub-network's miR-NAs, the TAM online tool was utilized. Using this tool, we identified miRNAs families for these sub-networks.
Hub miRNAs in TF-miRNA-TG_A and TF-miRNA-TG_B sub-networks are essential and crucial, as they regulate more genes of the subnetworks at a post-translational regulation level and can impact biological processes. Therefore, for each TF-miRNA-TG subnetwork, five high degree miRNAs were selected and reported as significant miRNAs . Based on the findings, hsa-miR-193b, hsa-miR-192, hsa-miR-215, hsa-miR-34a, and hsa-miR-16 were significant miRNAs for TF-miRNA-TG_A subnetwork, and hsa-miR-16, hsa-miR-92a, hsa-miR-30a, hsa-miR-7, and hsa-miR-26b were found to be significant miRNAs for TF-miRNA-TG_B subnetwork as well. Among these miRNAs, hsa-miR-16 regulates more genes in both subnetworks. According to the literature, some of these miRNAs have been studied in SARS coronavirus.
Kyung Hee Choi et al. 45 found that has-miR-193 is a dual-strand tumor suppressor and a novel therapeutic target for lung cancer. In another study, Huajun Hu et al. 46 reported that this miRNA is a tumor suppressor in Non-small cell lung cancer (NSCLC). Liang Sun et al. 47 concluded that hsa-miR-193b regulates the RAB22A oncogene, inhibits breast cancer growth, and may have significant implications for cancer therapy. In addition, this miRNA regulates breast cancer cell migration and vasculogenic mimicry by DDAH1 48 . Moreover, it could function as a tumor-suppressive miRNA in breast cancer 49 and inhibits breast cancer metastasis 50 .
Martyna Filipska et al. 51 found that hsa-miR-192-5p has a functional role in squamous cell lung cancer cells. In Peng Zou et al. 52 reported that this miRNA suppresses the progression of lung cancer bone metastasis by targeting TRIM44. Moreover, this miRNA has been introduced as a prognostic marker for NSCLC participants 53 . Importantly, hsa-miR-192 induces Cisplatin-resistance, inhibits cell apoptosis in lung cancer 54 and the proliferation, migration, and invasion of osteosarcoma cells, and promotes apoptosis 55 . Xiaopan et al. 56 reported that hsa-miR-215 suppresses proliferation and migration of non-small cell lung cancer cells(NSCLC). This miRNA is downregulated in NSCLC tissues and may play a key role in the development of NSCLC. The lower expression of has-miR-215 in NSCLC is negatively associated with lymphatic metastasis and TNM staging 57 . This miRNA targets ZEB2 in human non-small cell lung cancer and functions as a tumor suppressor 58 . Ariana Centa et al. 59 expressed that has-miR-34a-5p is identified as the regulator of mRNA targets involved in endothelial, inflammatory signaling pathways, and viral diseases. Furthermore, in the present study, the expression of this miRNA was significantly down-regulated in the COVID-19 patients compared to the Controls. Also, Martin Hart et al. 60 , in their systems biology, analysis identified miR-34a as strongly associated with pathogenesis. In another study, Rieko Aida et al. 61 reported that apigenin might induce apoptosis by down-regulating SNAI1 through miR-34a-5p up-regulation in A549 cells. Woo Ryung Kim et al. 62 reported that hsa-miR-16-5p is commonly bound to SARS-CoV, MERS-CoV, and SARS-CoV-2. In Zofia Wicik et al. 63 showed that this miRNA could regulate ACE2 networks. Moreover, this miRNA can link the pathogenesis of HIV-1 and malaria [64][65][66] . Similarly, Jianghong Wei et al. 67 found that overexpression of miR-16 inhibited the growth and metastasis of the DMS-53 lung cancer cells. Alireza Paniri et al. 68 reported that hsa-miR-26b-5p strongly targets ACE2 and have an important effect on SARS coronavirus. Like has-miR-16-5p, has-mir-26b-5p can regulate ACE2 networks as well 63 . Moreover, this paper reported that has-miR-26b-5p may plays a significant role in the pathogenesis of HF in COVID-19 patients. The effect of this miRNA in SARS coronavirus was studied by Laura Teodori and her colleagues 69 . Moreover, Yang Gao et al. 70 reported that this miRNA plays an important role in tumor suppression in lung cancer. According to M Xia et al., this miRNA could suppress lung cancer cells' proliferation, migration, and invasion. Min Jiang et al. 71 reported that has-miR-92a family could be ideal biomarkers for cancer diagnosis and prognosis. Also, our study revealed that the expression of has-mir-92a was upregulated in lung squamous cell carcinoma (LUSC). Besides, this miRNA could promote growth, metastasis, and chemoresistance in NSCLC cells 72 . This miRNA was thus introduced as a plasma biomarker for small cell lung cancer 73 . Jianhua Gong et al. 74 revealed that the has-miR-92a up-regulation could significantly induce proliferation and inhibit apoptosis of lung cancer cells. Jianjie Zhu et al. 75 revealed that the upregulation of has-mir-30-5p in lung cancer cell lines inhibited cell proliferation in vitro and in vivo. This miRNA suppresses lung cancer progression by targeting SIRT1 76 . Also, the lack of its expression promotes the growth of lung cancer cells by targeting MEF2D. Moreover, Xiaowei Quan et al. 77 and Ruixue Tang et al. 78 revealed that miR-30a-5p expression is downregulated in NSCLC. In addition, the increase in miR-30a-5p level could enhance Bax protein level and decrease Bcl-2 protein level 77 . In the field of pharmaceutical research, Xiaojie Xu et al. 79 reported that miR-30a-5p enhances paclitaxel sensitivity in non-small cell lung cancer through targeting BCL-2 expression. Haiping Xiao et al. 80 believed that has-miR-7-5p suppresses tumor metastasis of NSCLC by targeting NOVA2. Plus, Kenneth Lundstrom 81 revealed that Rotavirus (RV) miR-7 can inhibit rotavirus replication by targeting the RV nonstructural protein 5. In another study, it was found that has-miR-7 could repress fibrogenesis of lung fibroblasts induced by TGF-β1 82 . In addition, Xiaofei Zhang et al. 83 reported that the overexpressed CDR1as functions as an oncogene to promote the tumor progression via miR-7 in non-small-cell lung cancer.
From the perspective of pharmacological studies, our finding shows that 470 drugs target TF and non-TF genes in both TF-miRNA-TG_A and TF-miRNA-TG_B subnetworks. Of 470 drugs, 62 drugs target both TF and non-TF genes, 95 drugs target non-TF genes, and 436 drugs target TF genes. From among the 470 obtained drugs, the drugs which target more genes were selected and discussed. In the drug-gene network, the drugs with a degree of 3 or above were selected and reported as potential and effective drugs for treating patients infected with SARS coronavirus. These drugs, including FLUOROURACIL, EPIRUBICIN, and FOSTAMATINIB www.nature.com/scientificreports/ at least three genes in the drug-gene network. Also, the drugs with a degree of 4 or above were selected and reported for the drug-TF network. These high-degree drugs were CISPLATIN, SIROLIMUS, CYCLOPHOSPHA-MIDE, METHYLDOPA, VORINOSTAT, OLAPARIB, MITOXANTRONE, FLUOROURACIL, EVEROLIMUS,  PACLITAXEL, DAUNORUBICIN, ZINC CHLORIDE, METHOTREXATE, CARBOPLATIN, BORTEZOMIB, NICLOSAMIDE, and ETOPOSIDE. Of the reported drugs, CISPLATIN targets 11 Transcription factor genes and may have a crucial impact on SARS coronavirus disease. We found that some of these drugs have been studied in SARS-CoV and SARS-CoV-2 coronaviruses, and others can be assumed as candidate drugs for SARS-CoV-2 coronavirus disease therapeutic. Shamim I. Ahmad, in his recent study, revealed that FLUOROURACIL, in combination with deoxyribose and deoxyribonucleosides, can be a therapeutic option for SARS coronavirus 84 . EPIRUBICIN, VAPREOTIDA, and SAQUINAVIR have been proposed as key drugs in SARS coronavirus treatment 85 . Also, Strich et al. 86 introduced the FOSTAMATINIB as a potential therapeutic for COVID-19. Moreover, FOSTAMATINIB has the potential to treat serious outcomes of coronavirus COVID-19, including acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) 87,88 . In addition, several studies evaluated and showed the impact of FOSTAMATINIB on SARS coronavirus [89][90][91][92] . The mTOR signaling plays a crucial role in MERS-CoV infection 93 . In this regard, Yadi Zhou et al. 34 observed that the SIROLIMUS is an inhibitor of mTOR with both antifungal and antineoplastic properties. In addition, this drug has been presented as a viral protein expression blocker 94 . Swaroop Revannasiddaiaha et al. 95 showed that CYCLOPHOSPHAMIDE had a potential role in mitigation of acute respiratory distress syndrome among patients with SARS-CoV-2. Moreover, Brocato et al. 96 , Othenin-Girard et al. 97 , Corso et al. 98 , Schaecher et al. 99 , and Revannasiddaiah et al. 95 evaluated and showed the impact of CYCLOPHOSPHAMIDE on SARS coronavirus disease with different approaches.
Al-Rashedi et al. 100 noted that the OLAPARIB is a potential drug for treating patients infected with SARS-COV-2. MITOXANTRONE has also been introduced as potential inhibitors of SARS-CoV-2 M pro101 . Safavi et al. 102 showed that the METHOTREXATE silence the immune activation in patients with COVID-19. Also, Sujoy Khan et al. 103 proposed this drug as a potential drug for treating patients infected with COVID-19. Additionally, this drug has a protective effect on SARS-CoV-2 infection via downregulating ACE2 104 . In this study, another drug that we have reported as a potential drug for SARS coronavirus was NICLOSAMIDE. This drug has previously been reported as an antiviral agent against COVID-19 105,106 . Other studies have reported the potential of this drug in treating patients infected with COVID-19 107,108 . Different studies have been undertaken on the effect of ETOPOSIDE on SARS coronavirus disease [109][110][111][112] . In our previous research based on protein-protein-network analysis, we proposed PACLITAXEL, CARBOPLATIN, BORTEZOMIB, VORINOSTAT, and DAUNORUBICIN as potential drugs for SARS-CoV-2 coronavirus treatment 21 . In this study, PACLITAXEL was introduced as the most potent therapeutic candidate drug. In previous research, rare studies examined the effect of this drug on SARS-CoV-2 disease.

Conclusion
In this study, focusing on the gene expression profile of SARS-CoV samples, an attempt was made to identify effective drugs for the treatment of this disease with a gene co-expression network-based approach. Given that the genomes of SARS-CoV and SARS-CoV-2 are very similar, it is expected that the drugs introduced to treat SARS-CoV coronavirus would also be effective in treating SARS-CoV-2 disease. Current research aimed to discover novel potential drugs for SARS-CoV disease in order to treating SARS-CoV-2 coronavirus based on a co-expression network analysis. To this end, at first, significant DEGs in normal and SARS-CoV infected samples were selected and then the gene co-expression network was reconstructed and two gene modules were discovered as significant modules. Then, two significant gene modules were discovered from the reconstructed co-expression network. Next, for the obtained modules, two sub-networks named TF-miRNA-TG_A and TF-miRNA-TG_B were drawn. Afterward, the list of the drugs targeting TF-miRNA-TG_A and TF-miRNA-TG_B sub-networks' genes was extracted, and two drug-gene and drug-TF interaction networks were drawn. Eventually, five drugs including FLUOROURACIL, CISPLATIN, SIROLIMUS, CYCLOPHOSPHAMIDE, and METHYLDOPA are proposed as poteintial drugs for SARS-CoV-2 coronavirus treatment. As well as, ten miRNAs including miR-193b, miR-192, miR-215, miR-34a, miR-16, miR-16, miR-92a, miR-30a, miR-7, and miR-26b were found to be significant miRNAs in treating SARS-CoV-2 coronavirus.

Methods
Dataset and preprocessing. The gene expression data used in this work were downloaded from the NCBI Gene Expression Omnibus (GEO) database with the accession number GSE1739. This data contains gene expression profiles of normal and Severe Acute Respiratory Syndrome (SARS) infected patients' blood samples. To assign probes to gene IDs, the annotation file published by Affymetrix was used.
In this article, a network-based approach was applied to discover therapeutic drugs for SARS-CoV-2 disease. To this end, at first, differentially expressed significant genes (p_value < 0.05) in healthy and SARS-CoV infected samples were selected, and then the gene co-expression network was reconstructed in STRING database for the filtered genes. www.nature.com/scientificreports/ Network reconstruction and module extraction. At first, differentially expressed genes were extracted for the normal and SARS infected groups. In order to calculate the differentially expressed genes, adjusted p_ value was calculated using Benjamini & Hochberg false discovery rate method. Then, 1441 genes with p-values less than 0.05 were assumed as primary genes, which were then used in network reconstruction. The list of mentioned primary genes is reported in Supplementary file S1. Afterward, the gene co-expression network is reconstructed by primary genes in STRING database. In this web tool, the minimum required interaction score parameter is adjusted to 0.04. In order to analyze the reconstructed co-expression network, the Cytoscape software 43 version 3.8.2 was used. ClusterViz 113 plugin was used to identify gene modules (highly interconnected regions) in the co-expression network. ClusterViz is a Cytoscpae plugin, which discovers modules in a biological network using various clustering algorithms. This plugin contains three commonly used clustering algorithms, including FAG-EC, EAGLE, and MCODE.
In this research, we have applied all of the algorithms and the results did not have significant difference, so we have decided to select one of them. Therefore, the MCODE (Molecular Complex Detection) algorithm was used to find the gene co-expression modules. MCODE is a graph theoretic clustering algorithm for discovering strongly connected regions in a given network 113 . This algorithm selects the seed nodes and expand them based on the density of the cluster and density of the local neighborhood 113 . The MCODE algorithm was performed with the following parameters: Degree threshold = 2, NodeScore Threshold = 0.2, K-Core Threshold = 2, and Maxdepth = 100.
Transcription factors, miRNAs, and target genes interaction network. To investigate the effect of Transcription Factors (TF) and microRNAs (miRNA) on target genes (TG), the TF-miRNA-TG sub-networks were reconstructed for gene modules.
Transcription factors (TF) are proteins that regulate the rate of transcription of genetic information from DNA to messenger RNA 114 . miRNAs are small non-coding RNAs that function in RNA silencing and post-transcriptional gene regulation 23,115 . Both TFs and miRNAs regulate gene expression 116 . To get regulatory interactions information of TFs-TGs, the TRRUST online database was utilized. This database contains 8,444 and 6,552 TF-target regulatory relationships of 800 human TFs and 828 mouse TFs, respectively. In addition to TFs-TGs regulatory information, TFs-miRNAs regulatory interactions are essential. This information is obtained from the TransmiR V2 42 database. TransmiR v2.0 incorporates 3,730 TF-miRNA regulatory interactions, covering 623 TFs and 785 miRNAs for 19 organisms. In this study, the information of miRNAs-TGs regulatory interactions was retrieved from the miRWalk 2.0 database. This database contains both validated and predicted interactions. In this study, only experimentally validated miRNAs-TGs interactions were considered.
In order to analyze the reconstructed co-expression network, the Cytoscape software was used. After analyzing the network, generally, 391 nodes and 1273 edges were observed in the reconstructed co-expression network. ClusterViz 113 plugin was used to identify gene modules (highly interconnected regions) in the co-expression network. ClusterViz is a Cytoscpae plugin, which discovers modules in a biological network using various clustering algorithms. This plugin contains three commonly used clustering algorithms, including FAG-EC, EAGLE, and MCODE. The MCODE (Molecular Complex Detection) algorithm was used to find the gene co-expression modules in this study. The MCODE algorithm was performed with the following parameters: Degree threshold = 2, NodeScore Threshold = 0.2, K-Core Threshold = 2, and Maxdepth = 100. After clustering the gene co-expression network, we discovered two significant modules (Module A and Module B) in the co-expression network. Enrichment analysis. The Database for Annotation, Visualization, and Integrated Discovery (DAVID ) v6. 8 117,118 was used for the enrichment analysis of the genes. The gene ontology and pathway enrichment analysis were done for obtained gene modules.
To identify miRNAs family, the TAM online tool 119 was applied. To do this, all the miRNAs in TF-miRNA-TG sub-networks were imported to the TAM tool separately, and then significant miRNA families were identified.

Drug-Gene interaction network.
To identify the drugs that target TF-miRNA-TG_A and TF-miRNA-TG_B TF and non-TF genes, the Drug Gene Interaction Database (DGIdb) 41 was used. This database retrieves drug-gene interaction information from 24 other related databases. In this study, to identify drug-genes interaction, only approved drugs were used. GSEA and candidate drugs validation. In order to evaluate the proposed drugs for SARS-CoV-2 disease, the GSEA was performed by querying the Enrichr database 120 . To this end, the Enrichr database was utilized to perform the Connectivity Map(CMAP) analysis 121  www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.