key: cord-1043040-x7bejjkk authors: Martin, Roman; Löchel, Hannah F.; Welzel, Marius; Hattab, Georges; Hauschild, Anne-Christin; Heider, Dominik title: CORDITE: the curated CORona Drug InTERactions database for SARS-CoV-2 date: 2020-06-20 journal: iScience DOI: 10.1016/j.isci.2020.101297 sha: 52f3259b1cecffb13c8ec572b113744880f2194b doc_id: 1043040 cord_uid: x7bejjkk Summary Since the outbreak in 2019, researchers are trying to find effective drugs against the SARS-CoV-2 virus based on de novo drug design and drug repurposing. While the former approach is very time-consuming and needs extensive testing in humans, drug repurposing is more promising, as the drugs have already been tested for side-effects, etc. Currently, there is no treatment for COVID-19 that is clinically effective, but there is a huge amount of data from studies that analyze potential drugs. We developed CORDITE to efficiently combine state-of-the-art knowledge on potential drugs and make it accessible to scientists and clinicians. The web interface also provides access to an easy-to-use API that allows a wide use for other software and applications, e.g., for meta-analysis, design of new clinical studies, or simple literature search. CORDITE is currently empowering many scientists across all continents and accelerates research in the knowledge domains of virology and drug design. Availability: CORDITE is available athttps://cordite.mathematik.uni-marburg.de Coronaviruses can cause major outbreaks with fatal pneumonia, e.g., the severe acute respiratory syndrome coronavirus (SARS-CoV) in 2003 with a fatality rate of around 10% (Lee et al., 2003) or the Middle East Respiratory Syndrome Coronavirus (MERS-CoV) in 2012 with a fatality rate of 35% (de Groot et al., 2013) . In 2019, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), caused the disease COVID-19 and led to a pandemic (Zhu et al., 2020) . The common symptoms of COVID-19 include fever, cough, shortness of breath, and dyspnea. A more severe progression of the COVID-19 disease may lead to pneumonia and severe acute respiratory syndromes. Unfortunately, there are no specific drugs or vaccines approved for the treatment of COVID-19, yet (Huang et al., 2020) . Traditionally, de novo drug design and drug repurposing are the common strategies scientists use to combat viral diseases. However, designing a new drug is very time-consuming and needs extensive testing before clinical use in humans. In contrast, the repurposing of approved drugs or drugs in clinical trials is more promising, in particular since these have already been tested for side-effects, toxicity, bioavailability, etc. At the moment, there is a huge amount of data from studies that analyze potential drugs based on computer simulations, in vitro studies, case studies, or clinical trials. However, none of the proposed treatments for COVID-19 have been proven to be clinically effective, yet. Potential drugs can target viral proteins or human proteins. At the time of writing, the most important viral target molecules that have been investigated for drug repurposing are: SARS-CoV-2 spike protein, main protease (3CLpro, Mpro), papain-like proteinase (PLpro), and RNAdependent RNA polymerase (RdRp) (Out et al., 2020; Zhang et al., 2020) . While the main targets on the human cell membrane are: Angiotensin-converting Enzyme 2 (ACE2) and Transmembrane protease serine 2 (TMPRSS2) (Aronson and Ferner, 2020) . Besides these targets, several other candidates have been investigated as potential targets. Such investigations occurred on both viral and human sides, e.g., the methyltransferase and furin, respectively (Bestle et al., 2020; Khan et al., 2020) . Also, several potential drugs were announced at the beginning of the pandemic, which later proved to be insufficiently effective or even harmful to patients, such as hydroxychloroquine or chloroquine phosphate. For instance, Gautret et al. (2020) reported a significant decrease in viral load in patients treated with hydroxychloroquine compared to controls. However, others have shown that these potential drugs can have severe side effects such as cardiac toxicity and also increased mortality (Mehra et al., 2020) . Nevertheless, some of those studies showing higher mortality were later retracted because of a lack of transparency concerning the data (Mehra et al., 2020b) . Other potential drug candidates, such as a combination of Lopinavir with Ritonavir (also known as Kaletra), was first considered a promising candidate (Jeon et al., 2020) but later shown to be ineffective for SARS-CoV-2 (Cao et al., 2020) . Nevertheless, many new drug candidates have been proposed by using different computational approaches, which address different viral and human targets It is expressed as a membrane-bound protein in the lungs, heart, kidneys, stomach, spleen, intestine, bone marrow, kidney, liver, brain, testis, and placenta (Shenoy et al., 2011; Donoghue et al., 2020; Soler et al., 2020) . Moreover, Hoffmann et al. (2020) showed that SARS-CoV-2 depends on TMPRSS2, which therefore can serve as a target in therapy as well. TMPRSS2 primes the spike protein of SARS-Cov-2 to enable cell-entry on ACE2. The authors further suggested that ACE-2 binding to the spike protein might have a higher affinity compared to SARS. This may explain the infection of the upper respiratory system by SARS-CoV-2, which has not been observed in SARS, since in the upper respiratory system less ACE2 is expressed. Structurally, each virion is approximately 50-200 nm in diameter (Chen et al., 2020) . Like other coronaviruses, SARS-CoV-2 has four structural polyproteins, known as the S (spike), E (envelope), M (membrane), and N (nucleocapsid) proteins. The latter, N protein holds the RNA genome, while the S, E, and M proteins together are part of the viral envelope. Upon successful attachment of the virus to the membrane of the host cell via the spike protein, the virus fuses with the membrane (Wu et al., 2020 , Hoffmann et al., 2020 . Next, the endosome sees a drop of pH and the genomic RNAss(+) is released into the cell's cytoplasm (Yang et al., 2004) . As shown in Figure 1 The knowledge from the life cycle of the SARS-CoV-2 virus offers a great potential to find suitable drugs for the already known targets. To combine the large amounts of current knowledge, we centralize the aggregated information from published articles and preprints about potential drugs, targets, and their interactions, into CORDITE (CORona Drug InTEractions database), which includes a web interface and an API. CORDITE integrates many functionalities to enable users to access, sort, and download relevant data to conduct meta-analyses, to design new clinical trials, or even to conduct a curated literature search. CORDITE automatically incorporates publications from PubMed (https://www.ncbi.nlm.nih.gov/pubmed/), bioRxiv (https://www.biorxiv.org/), chemRxiv (https://www.chemrxiv.org/), and medRxiv (https://www.medrxiv.org/) that report information on computational, in vitro, or case studies on potential drugs for COVID-19. Besides original research, we also included reviews and comments in the database. The information from the articles and preprints are manually curated by moderators and can be accessed via the web interface or the open API. Users can directly access the publications, interactions, drugs, targets, and clinical trials. The interactions and the effectiveness of certain drugs are also shown as positive/negative results in a transparent way. Thus, the user can easily see the contradictory results of different studies and can take these into account for, e.g., meta-analyses. This access may offer easier integration for future software or apps that may need curated data. Moreover, we include registered clinical trials from the National Institute of Health or NIH (https://clinicaltrials.gov/) for COVID-19. The CORDITE database is updated through manual curation. The current summary for interactions, drugs, etc., is reported in Table 1 . An introductory video for the use of CORDITE can be found on the website. CORDITE provides an easy-to-use web interface and also access to an easy-to-use API that allows a wide use for other software and applications, e.g., for meta-analysis, design of new clinical studies, or simple literature search. CORDITE is currently the largest curated database for drug interactions for SARS-CoV-2 and therefore it offers the opportunity to accelerate research in the knowledge domains of virology and drug design. Because of the popularity and impact of the CORDITE website and database, we plan to continue hosting and managing CORDITE throughout the COVID-19 pandemic. Moreover, we will continuously improve the website and database to establish a standard tool for future diseases. We believe our efforts are essential to help researchers and clinicians to get an overview of existing studies and to accelerate the global search for new treatments. CORDITE aggregates available knowledge on drugs for COVID-19, however, in particular the preprints have not been peer-reviewed. Thus, CORDITE cannot guarantee the reliability of the studies and their results, in particularly the effectiveness of the drugs. The authors thank Hagen Dreßler for technical assistance. RM developed and implemented the database and the API. HFL developed and implemented the web interface and curated approximately 2.3% of the data. MW developed the automated literature crawler and curated approximately 11.3% of the data. GH created figure 1 and the visual identity of the web interface and contributed to the initial draft. ACH contributed to the project design, created figure 2, and curated approximately 7.4% of the data. DH designed and supervised the project, wrote the manuscript, and curated approximately 79.0% of the data. All authors revised the final manuscript. The authors declare no competing interests. Interactions 849 Drugs 639 Publications 324 Clinical trials 247 Further information and requests for resources should be directed to and will be fulfilled by the Lead Contact, Dominik Heider (dominik.heider@uni-marburg.de). This study did not generate new unique data. This study did not generate data. The curated data can be accessed freely via the API. The data integrated with CORDITE is aggregated in three procedural steps: (A) automated literature aggregation, (B) manual database curation, and (C) the web interface enabling users to search and export subsets of the curated data. The complete diagram depicting these steps is shown in Figure 2 . Literature aggregation is carried out using a crawler written in Python 3.7.7. The PubMed (https://www.ncbi.nlm.nih.gov/pubmed/) entries are accessed using Biopython 1.76 via requests to the Entrez meta-search engine. Access of chemRxiv (https://www.chemrxiv.org/) entries is carried out using the figshare API (https://figshare.com/). Literature available on medRxiv (https://www.medrxiv.org/) and bioRxiv (https://www.biorxiv.org/) is aggregated by a feed parser. Information regarding ongoing clinical trials is fetched by URL requests. The aggregated data is filtered for relevant publications / clinical trials. Moreover, suggested papers received via email are added to the database as well and curated. The list of automatically aggregated entries from different resources is manually curated and appropriate clinical trials and publications are selected. For each publication and clinical trial, we extract the proposed drugs, drug-viral-target, or drug-host-target interactions and integrate them into the CORDITE database. In particular, we store evidence of both confirmed and disproven interactions and indicate them as positive or negative. CORDITE is a relational database implemented in MariaDB 10.1 consisting of eight tables and fourteen views. The publications, clinical trials, interactions, and molecules, including drugs and targets, are stored in individual tables, while the other tables implement cross-references between the content tables. The corresponding API to read the database was implemented in PHP 7.2. New data can easily be inserted, modified, and deleted via a hidden editorial website written in PHP and Bootstrap 4. The editorial website further ensures the quality of the content by performing additional integrity checks and removing duplicates. The CORDITE web interface enables an extensive search for drugs, targets, interactions, publications, and clinical trials associated with COVID-19. The functionality allows the user to sort and select drugs or targets of interest and view all corresponding information. Moreover, it is possible to order drugs according to the number of publications. To this end, only peerreviewed research articles are taken into account. To ensure smooth user interface interactions, we use the JavaScript react v. 16 Biochemical characterization of a recombinant SARS coronavirus nsp12 RNA-dependent RNA polymerase capable of copying viral RNA templates Coronavirus main proteinase (3CLpro) structure: basis for design of anti-SARS drugs Drugs and the renin-angiotensin system in covid-19 TMPRSS2 and furin are both essential for proteolytic activation and spread of SARS-CoV-2 in human airway epithelial cells and provide promising drug targets A Trial of Lopinavir-Ritonavir in Adults Hospitalized with Severe Covid-19 Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study SARS coronavirus papain-like protease inhibits the type I interferon signaling pathway through interaction with the STING-TRAF3-TBK1 complex Middle East respiratory syndrome coronavirus (MERS-CoV): Announcement of the Coronavirus Study Group A novel angiotensinconverting enzyme-related carboxypeptidase (ACE2) converts angiotensin I to angiotensin 1-9 Hydroxychloroquine and azithromycin as a treatment of COVID-19: results of an openlabel non-randomized clinical trial Return of the Coronavirus: 2019-nCoV The origin, transmission and clinical therapies on coronavirus disease 2019 (COVID-19) outbreak-an update on the status Coronavirus disease 2019: What we know SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. The Lancet Computational Strategies to Combat COVID-19: Useful Tools to Accelerate SARS-CoV-2 and Coronavirus Research Identification of antiviral drug candidates against SARS-CoV-2 from FDA-approved drugs Targeting SARS-CoV-2: a systematic drug repurposing approach to identify promising inhibitors against 3C-like proteinase and 2'-O-ribose methyltransferase A major outbreak of severe acute respiratory syndrome in Hong Kong Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis Retraction-Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis. The Lancet Characterization of spike glycoprotein of SARS-CoV-2 on virus entry and its immune cross-reactivity with SARS-CoV Exploring the SARS-CoV-2 virus-hostdrug interactome for drug repurposing ACE2, a promising therapeutic target for pulmonary hypertension Pharmacologic modulation of ACE2 expression Receptor Recognition by the Novel Coronavirus from Wuhan: an Analysis Based on Decade-Long Structural Studies of SARS Coronavirus Analysis of therapeutic targets for SARS-CoV-2 and discovery of potential drugs by computational methods Evolution of the novel coronavirus from the ongoing Wuhan outbreak and modeling of its spike protein for risk of human transmission pH-dependent entry of severe acute respiratory syndrome coronavirus is mediated by the spike glycoprotein and enhanced by dendritic cell transfer through DC-SIGN Structural basis for inhibition of the RNA-dependent RNA polymerase from SARS-CoV-2 by remdesivir Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors A novel coronavirus from patients with pneumonia in China