Abstract
The rapid emergence of SARS-CoV-2 variants with multi-sites mutations is considered as a major obstacle for the development of drugs and vaccines. Although most of the functional proteins essential for SARS-CoV-2 have been determined, the understanding of the COVID-19 target-ligand interactions remains a key challenge. The old version of this COVID-19 docking server was built in 2020, and free and open to all users. Here, we present nCoVDock2, a new docking server to predict the binding modes for targets from SARS-CoV-2. First, the new server supports more targets. We replaced the modeled structures with newly resolved structures and added more potential targets of COVID-19, especially for the variants. Second, for small molecule docking, Autodock Vina was upgraded to the latest version 1.2.0, and a new scoring function was added for peptide or antibody docking. Third, the input interface and molecular visualization were updated for a better user experience. The web server, together with an extensive help and tutorial, are freely available at: https://ncovdock2.schanglab.org.cn.
Graphical Abstract
Graphical Abstract.
INTRODUCTION
It has been over three years since the World Health Organization (WHO) announced coronavirus disease 2019 (COVID-19) as a great pandemic on March 11, 2020. As of 26 February 2023, in total over 758 million cases have been confirmed and 6.8 million deaths have been reported globally. COVID-19 is caused by a newly identified β type of coronavirus, sharing highly genetic homology to SARS-CoV (severe acute respiratory syndrome coronavirus) (1,2). The pathogen was officially named as SARS-CoV-2 by the International Committee on Taxonomy of Viruses (ICTV) and WHO. It typically leads to acute respiratory illness and symptoms such as cough, fever, headache, nausea, vomiting and diarrhea (3). In most people, the symptoms are mild and tolerable, while elders or people with health issues may face much more severe situations including multi-organ dysfunction and acute respiratory distress syndrome (ARDS). Currently, drugs such as Paxlovid, Xocova and Molnupiravir, and vaccines including mRNA-based and inactivated vaccines, are approved for COVID-19 and provide major clinical benefits for patients (4–8).
However, as a single-stranded RNA virus, SARS-CoV-2 presents a strong trend to evolve numerous variants with multiple mutations in the whole genome, with the potential to escape from the vaccines or be resistant to the drugs (9–11). WHO has announced five variants of concerns (VOCs) including Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2) and Omicron (B.1.1.529) as well as eight variants of interest (VOIs) such as Epsilon (B.1.427 and B.1.429), Eta (B.1.525), Kappa (B.1.617.1), Mu (B.1.621), Lambda (C.37), Theta (P.3), Zeta (P.2) and Lota (B.1.526) (https://viralzone.expasy.org/9556). Most of the mutations are located on the Spike protein, especially in the receptor binding domain (RBD). As for dominant variant Omicron, about 30 mutations occur in the S protein RBD region, which directly facilitate the virus entry and immune evasion, and finally increase the infection (12,13). In addition, mutations on Nsp1, Nsp3, Nsp4, Nsp5 (3CL protease), Nsp6, Nsp12, Nsp13, Nsp14 and Nsp15 are also identified. Understanding the impact of mutations in three dimensional structures of virus proteins, as well as molecular recognition and interactions, are of great interest in the study of molecular mechanisms of the viral life cycle, and may also provide clues for drug and vaccine design.
In previous studies, some servers have been developed for the prediction of COVID-19 targets. The web server of Virus-CKB applies pharmacology target mapping to rapidly predict the FDA-approved drugs for COVID-19 (14). DINC-COVID is a webserver that performs the simultaneous docking of a ligand against multiple receptor conformations (15). The web server of MolAICal supplies a deep learning model for generating new compounds in the 3D pocket of the SARS-CoV-2 Main protease (16). DockThor-VS is a virtual screening platform to repurpose known drugs against six selected proteins of SARS-CoV-2 (17). The web server of D3Targets-2019-nCoV provides structure-based (D3Docking) and ligand based (D3Similarity) approaches for target prediction and virtual screening (18,19). Unlike these servers, the nCoVDock2 server predicts the binding modes between COVID-19 targets and various types of ligands, not only including small molecules, but also peptides and antibodies. In the recent CASP15 challenge, our group (CoDock) ranked No. 1 in the ligand prediction category. In the CASP14-CAPRI protein assembly prediction challenge, our group (Chang) ranked No. 1 in the scoring experiment and No. 3 in the prediction experiment (20). These results confirm the robust performance of our docking protocols. In this new server, we have updated the target structures of SARS-CoV-2 and improved the docking method, with the aim to provide a tool to better understand targets-ligand interactions for COVID-19.
MATERIALS AND METHODS
Targets of SARS-CoV-2
There are totally 2964 experimental resolved structures of SARS-CoV-2 proteins available in the Protein Data Bank till 12 March 2023. Most of the proteins encoded by the viral genome have three dimensional structures determined, providing fundamental information for the functional study and the development of drug or vaccine. In the new version of the COVID docking server, new structures have been added into the target list, including ORF7a, ORF9b, nucleocapsid protein (full length), as well as multiple variants for the S protein. The full length and receptor binding region (RBD) of the S protein for different variants, such as Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2), Omicron (B.1.1.529), Kappa (B.1.617.1) and Epsilon (B.1.427 and B.1.429) are provided for investigation. The N-terminal domain (NTD) of the S protein is recently found to involve in antibody evading and structures of NTD are also provided as new targets (21). The modeled structures are replaced by the newly available experimental structures, such as nsp1, nsp2, nsp13, nsp14, ORF3A and ORF8. The structures of nsp3, Papain-like protease (PLpro) and nsp16/10 are updated to 5RSF, 7RZC and 6W4H, respectively, with lower resolution values and higher quality (22,23). The structure of Main protease (Mpro) is updated to 7SI9, the room-temperature X-ray structure in complex with the approved drug, PF-07321332 (24). The detailed information of targets and structures is available on the target annotation webpage (see Supplementary Text and Supplementary Table S1 for details).
Docking methods
Autodock Vina (25) is used as the docking engine for small molecule docking. In the new server, Autodock Vina is upgraded to the latest version 1.2.0 (26). The docking box is defined as the center of native ligand coordinates with dimensions of 30 Å × 30 Å × 30 Å to include residues of the entire cavity. Users can set exhaustiveness value options for the docking precision. The docking accuracy of Autodock Vina has been validated in earlier works (27). Except for the Autodock Vina score, we also provide a machine learning scoring function RF-Score as a complement to evaluate the binding affinities for small molecules (28). By testing the all 102 targets from the DUD-E dataset (26), the latest version Vina 1.2.0 program has an average EF 1% of 9.7 and better than that of 7.6 for Vina 1.1.2. Furthermore, the RF-Score substantially improves the virtual screening performance of Vina according to evaluations on the DUD-E dataset and the DEKOIS benchmark (29).
As a docking engine of peptide or antibody, the CoDockPP program provides a multistage fast Fourier transform (FFT)-based strategy for both the global docking and the site-specific docking (30). In our previous study, we have tested the CoDockPP on the Protein-Protein Docking Benchmark 5.0. Similar to the small molecule docking, a machine learning scoring function is added for peptide and antibody docking in the new server. The scoring function is based on the deep neural network, and transforms the atomic densities of each protein into SE(3)-equivariant representations (31). This new scoring model is trained on a set of decoy conformations generated from 851 nonredundant protein-protein complexes and outperforms the state of the art on data from the Protein-Protein Docking Benchmark 4.0.
Web implementation
The frontend interface utilizes the latest technologies in HTML5, CSS and JavaScript. The encapsulated jQuery library is used to realize the event handling and AJAX technology. The interactive NGL Viewers are applied for the visualization and analysis of the final docking results. It supports many file formats, including pdb, sdf and mol2, as well as many molecular representations, such as Cartoon, Surface and Licorice.
The backend API is implemented in php using the Thinkphp framework. A MySQL database is used to store the user's job data. The server is built on the Linux centos system and isolated by using the Docker container technology. When a docking job is submitted on the frontend, a unique job ID will be created for it. Then, the server processes it according to the current resource situation. After the job is completed, the docking result can be viewed according to the job ID in the frontend interface.
SERVER DESCRIPTION
Inputs of the small molecule docking
As shown in Figure 1A, users need to choose one of the COVID-19 targets and its wild type or mutants from the drop-down box. There are two computational modules of the small molecule docking. If ‘Docking’ is selected, only one small molecule should be uploaded, and the top 10 binding modes will be displayed on the result page. If ‘Batch Docking’ is selected, 10–20 molecules should be uploaded, and the top 10 molecules ranked by the scoring function will be displayed on the result page. The small molecules can be uploaded in smi, mol2 or sdf formats. Exhaustiveness is a parameter of the docking search. Increasing this value will increase the time linearly and increase the probability to find the minimum exponentially. An email will be sent to the user's email address, containing a directed link to access the docking results.
Figure 1.
Inputs and results for the small molecule docking in the single and batch modes. (A) Submission options of the small molecule docking. (B) Docking results of the single molecule docking. The crystal structure of the inhibitor is colored gray. The predicted top 1 pose is colored pink. (C) Docking results of the batch docking. The predicted binding poses of 10 inhibitors are shown in different colors.
Inputs of the peptide and antibody docking
Similar to small molecule docking, users need to choose the COVID-19 protein target firstly (see Figure 2A). Then, the peptide or antibody protein is uploaded in the pdb format. A PDB checker, such as Molprobity (32), should be used to check the file and fix any potential errors in the pdb file. The protein docking server can perform the global docking and site-specific docking to predict the binding mode between target and ligand proteins. In the text boxes, users can enter one constraint residue on the target interface and another one on the ligand interface. If only one residue on the ligand or target is defined, then the conformations with the specific residue on the interface of the complex are retained. When a user defines one residue on the receptor and the other one on the ligand simultaneously, then he needs to choose the constraint type: ambiguous constraint or multiple constraints. When the ambiguous constraint is selected, the conformations are retained with at least one selected residue on the interface. When the multiple constraints are selected, the conformations are retained with both of the residues on the interface.
Figure 2.
Inputs and results for the peptide and antibody docking. (A) Submission options of the peptide and antibody docking. (B) Docking results of the peptide docking. (C) Docking results of the antibody docking. The ligand structure from the original pdb is colored gray, and the predicted top 1 pose is colored pink.
Usage examples
To illustrate the usage and capabilities of the nCoVDock2 webserver, we present case studies showing how the webserver can be used to generate accurately docked models for the four docking types. To present the accuracy of the prediction, the RMSD values were calculated for these four cases. In general, the submitted ligand structures are not the crystal structures, so it is lack of reference experimental structures to calculate RMSD values. Therefore, the RMSD values are not shown on the result page of the webserver.
The Main protease (Mpro) and its approved drug, PF-07321332 are taken as an example for the small molecule docking. The Main protease is selected as the protein target. In the ‘Docking’ mode, the approved drug PF-07321332 in PDB 7SI9 is chosen as the input ligand. For the top rank binding modes, the predicted binding energy of the top 1 was -8.8 kcal/mol, the RF score was 7.0 pKd and the L_RMSD was 1.3 Å. As shown in Figure 1B, the predicted binding mode of the small molecule is very close to the crystal structure of the inhibitor. The Main protease and its 12 inhibitors are also taken as an example for the small molecule batch docking. As shown in Figure 1C, the top 10 small molecules are docked in the correct binding pocket.
The post-fusion state of the S2 segment of the spike protein is taken as an example for the peptide docking. A 5-helices structure extracted from the post-fusion state of 6-helices S2 (PDB ID: 7BZ5) is selected as the protein target. In the default global docking mode, the single helix was chosen as the ligand peptide. For the top rank binding modes, the scoring value of top 1 binding mode was ṇ611.2 kcal/mol, the CNN score is -9.2 and the L_RMSD was 0.9 Å. The spike protein and antibody are taken as an example for the antibody docking. The RBD domain of the spike protein (PDB ID: 6M0J) is selected as the target. The antibody structure in PDB 7BZ5 is chosen as a ligand, so it can be considered as an unbound global docking. For the top rank binding modes, the scoring value of top 1 binding mode was -440.9 kcal/mol, the CNN score is ṇ1.8 and the L_RMSD was 1.1 Å. As shown in Figures 2B and C, the best predicted conformations of peptide and antibody are both very similar to the native structures.
Summary of community usage statistics
We also analyzed the jobs on the COVID-19 server and made statistics data for the predictions (see Figure 3). The previous server has completed more than 30000 predictions. As shown in Figure 3A, the prediction jobs grow rapidly from 2020 to 2021. After 2022, the prediction jobs grow slowly, but the predictions for peptides and antibodies are increasing. As shown in Figure 3B, the batch docking and the peptide docking are the most commonly used modules. For the four job types, the most predicted targets are also different. For the small molecule docking and batch docking, the most predicted targets are Main protease, RdRp and Papain-like protease (see Figures 3C and D). For the peptide and antibody docking, the most docked targets are changed to the Spike protein, Membrane protein and ACE2 (see Figure 3E and F). These data are consistent with the current directions of drug design.
Figure 3.
Statistics of the predictions on the COVID-19 server. (A) The growth curve of prediction jobs. (B) Proportion of different docking types. (C–F) The top 5 targets for different docking types.
CONCLUSIONS
A new docking server, nCoVDock2 was constructed to predict the binding modes between the wild type and mutants of SARS-CoV-2 therapeutic targets and their potential ligands. The server provides a user-friendly interface and binding mode visualization for the predicted results, which makes it a useful tool for the drug discovery of COVID-19.
DATA AVAILABILITY
The nCoVDock2 web server, with the online documentation, is publicly available at https://ncovdock2.schanglab.org.cn.
Supplementary Material
ACKNOWLEDGEMENTS
The authors wish to thank Prof. Xiaoqin Zou of the University of Missouri for the docking protocol discussion. The authors also wish to thank Primary Biotech (www.pumeirui.com) for their assistance in designing the website.
Contributor Information
Kai Liu, Institute of Bioinformatics and Medical Engineering, School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou 213001, China.
Xufeng Lu, Institute of Bioinformatics and Medical Engineering, School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou 213001, China; Primary Biotechnology Co., Ltd., Changzhou 213125, China.
Hang Shi, Institute of Bioinformatics and Medical Engineering, School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou 213001, China.
Xiaojun Xu, Institute of Bioinformatics and Medical Engineering, School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou 213001, China.
Ren Kong, Institute of Bioinformatics and Medical Engineering, School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou 213001, China.
Shan Chang, Institute of Bioinformatics and Medical Engineering, School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou 213001, China.
SUPPLEMENTARY DATA
Supplementary Data are available at NAR Online.
FUNDING
National Natural Science Foundation of China [12074151]; Changzhou Sci. and Tech. Program [CE20205033]; Jiangsu University of Technology. Funding for open access charge: Jiangsu University of Technology.
Conflict of interest statement. None declared.
REFERENCES
- 1. Lu R., Zhao X., Li J., Niu P., Yang B., Wu H., Wang W., Song H., Huang B., Zhu N.et al.. Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding. Lancet. 2020; 395:565–574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Wu F., Zhao S., Yu B., Chen Y.M., Wang W., Song Z.G., Hu Y., Tao Z.W., Tian J.H., Pei Y.Y.et al.. A new coronavirus associated with human respiratory disease in China. Nature. 2020; 579:265–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Guan W.J., Ni Z.Y., Hu Y., Liang W.H., Ou C.Q., He J.X., Liu L., Shan H., Lei C.L., Hui D.S.C.et al.. Clinical characteristics of coronavirus disease 2019 in China. N. Engl. J. Med. 2020; 382:1708–1720. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Singh R.S.P., Toussi S.S., Hackman F., Chan P.L., Rao R., Allen R., Van Eyck L., Pawlak S., Kadar E.P., Clark F.et al.. Innovative randomized phase I study and dosing regimen selection to accelerate and inform pivotal COVID-19 trial of Nirmatrelvir. Clin. Pharmacol. Ther. 2022; 112:101–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Unoh Y., Uehara S., Nakahara K., Nobori H., Yamatsu Y., Yamamoto S., Maruyama Y., Taoda Y., Kasamatsu K., Suto T.et al.. Discovery of S-217622, a noncovalent oral SARS-CoV-2 3CL protease inhibitor clinical candidate for treating COVID-19. J. Med. Chem. 2022; 65:6499–6512. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Jayk Bernal A., Gomes da Silva M.M., Musungaie D.B., Kovalchuk E., Gonzalez A., Delos Reyes V., Martín-Quirós A., Caraco Y., Williams-Diaz A., Brown M.L.et al.. Molnupiravir for oral treatment of Covid-19 in nonhospitalized patients. N. Engl. J. Med. 2022; 386:509–520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Mulligan M.J., Lyke K.E., Kitchin N., Absalon J., Gurtman A., Lockhart S., Neuzil K., Raabe V., Bailey R., Swanson K.A.et al.. Phase I/II study of COVID-19 RNA vaccine BNT162b1 in adults. Nature. 2020; 586:589–593. [DOI] [PubMed] [Google Scholar]
- 8. Xia S., Duan K., Zhang Y., Zhao D., Zhang H., Xie Z., Li X., Peng C., Zhang Y., Zhang W.et al.. Effect of an inactivated vaccine against SARS-CoV-2 on safety and immunogenicity outcomes: interim analysis of 2 randomized clinical trials. JAMA. 2020; 324:951–960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Hemmer C.J., Löbermann M., Reisinger E.C.. [COVID-19: epidemiology and mutations : an update]. Radiologe. 2021; 61:880–887. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Lubin J.H., Zardecki C., Dolan E.M., Lu C., Shen Z., Dutta S., Westbrook J.D., Hudson B.P., Goodsell D.S., Williams J.K.et al.. Evolution of the SARS-CoV-2 proteome in three dimensions (3D) during the first 6 months of the COVID-19 pandemic. Proteins. 2022; 90:1054–1080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Pozzi C., Vanet A., Francesconi V., Tagliazucchi L., Tassone G., Venturelli A., Spyrakis F., Mazzorana M., Costi M.P., Tonelli M.. Antitarget, Anti-SARS-CoV-2 leads, drugs, and the Drug Discovery-Genetics Alliance perspective. J. Med. Chem. 2023; 66:3664–3702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Daria S., Bhuiyan M.A., Islam M.R.. Detection of highly muted coronavirus variant Omicron (B.1.1.529) is triggering the alarm for South Asian countries: associated risk factors and preventive actions. J. Med. Virol. 2022; 94:1267–1268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Fan Y., Li X., Zhang L., Wan S., Zhang L., Zhou F.. SARS-CoV-2 Omicron variant: recent progress and future perspectives. Signal Transduct. Target Ther. 2022; 7:141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Feng Z., Chen M., Liang T., Shen M., Chen H., Xie X.Q.. Virus-CKB: an integrated bioinformatics platform and analysis resource for COVID-19 research. Brief. Bioinf. 2021; 22:882–895. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Hall-Swan S., Devaurs D., Rigo M.M., Antunes D.A., Kavraki L.E., Zanatta G.. DINC-COVID: a webserver for ensemble docking with flexible SARS-CoV-2 proteins. Comput. Biol. Med. 2021; 139:15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Bai Q., Tan S., Xu T., Liu H., Huang J., Yao X.. MolAICal: a soft tool for 3D drug design of protein targets by artificial intelligence and classical algorithm. Brief. Bioinf. 2021; 22:bbaa161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Guedes I.A., Costa L.S.C., Santos K.B., Karl A.L.M., Rocha G.K., Teixeira I.M., Galheigo M.M., Medeiros V., Krempser E., Custódio F.L.et al.. Drug design and repurposing with DockThor-VS web server focusing on SARS-CoV-2 therapeutic targets and their non-synonym variants. Sci. Rep. 2021; 11:5543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Shi Y., Zhang X., Mu K., Peng C., Zhu Z., Wang X., Yang Y., Xu Z., Zhu W.. D3Targets-2019-nCoV: a webserver for predicting drug targets and for multi-target and multi-site based virtual screening against COVID-19. Acta Pharm. Sin. B. 2020; 10:1239–1248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Yang Y., Zhu Z., Wang X., Zhang X., Mu K., Shi Y., Peng C., Xu Z., Zhu W.. Ligand-based approach for predicting drug targets and for virtual screening against COVID-19. Brief. Bioinf. 2021; 22:1053–1064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lensink M.F., Brysbaert G., Mauri T., Nadzirin N., Velankar S., Chaleil R.A.G., Clarence T., Bates P.A., Kong R., Liu B.et al.. Prediction of protein assemblies, the next frontier: the CASP14-CAPRI experiment. Proteins: Struct. Funct. Bioinf. 2021; 89:1800–1823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Rosa A., Pye V.E., Graham C., Muir L., Seow J., Ng K.W., Cook N.J., Rees-Spear C., Parker E., Santos M.S.et al.. SARS-CoV-2 can recruit a heme metabolite to evade antibody immunity. Sci. Adv. 2021; 7:eabg7607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Schuller M., Correy G.J., Gahbauer S., Fearon D., Wu T., Díaz R.E., Young I.D., Carvalho Martins L., Smith D.H., Schulze-Gahmen U.et al.. Fragment binding to the Nsp3 macrodomain of SARS-CoV-2 identified through crystallographic screening and computational docking. Sci. Adv. 2021; 7:eabf8711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Rosas-Lemus M., Minasov G., Shuvalova L., Inniss N.L., Kiryukhina O., Brunzelle J., Satchell K.J.F.. High-resolution structures of the SARS-CoV-2 2'-O-methyltransferase reveal strategies for structure-based inhibitor design. Sci. Signal. 2020; 13:eabe1202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kneller D.W., Li H., Phillips G., Weiss K.L., Zhang Q., Arnould M.A., Jonsson C.B., Surendranathan S., Parvathareddy J., Blakeley M.P.et al.. Covalent narlaprevir- and boceprevir-derived hybrid inhibitors of SARS-CoV-2 main protease. Nat. Commun. 2022; 13:2268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Trott O., Olson A.J.. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 2010; 31:455–461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Eberhardt J., Santos-Martins D., Tillack A.F., Forli S.. AutoDock Vina 1.2.0: new docking methods, expanded force field, and Python bindings. J. Chem. Inf. Model. 2021; 61:3891–3898. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Kong R., Yang G., Xue R., Liu M., Wang F., Hu J., Guo X., Chang S.. COVID-19 docking server: a meta server for docking small molecules, peptides and antibodies against potential targets of COVID-19. Bioinformatics. 2020; 36:5109–5111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Li H., Leung K.S., Wong M.H., Ballester P.J.. Correcting the impact of docking pose generation error on binding affinity prediction. BMC Bioinf. 2016; 17:308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Wójcikowski M., Ballester P.J., Siedlecki P.. Performance of machine-learning scoring functions in structure-based virtual screening. Sci. Rep. 2017; 7:46710. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Kong R., Wang F., Zhang J., Wang F.F., Chang S.. CoDockPP: a multistage approach for global and site-specific protein–Protein docking. J. Chem. Inf. Model. 2019; 59:3556–3564. [DOI] [PubMed] [Google Scholar]
- 31. Derevyanko G., Lamoureux G.. Protein-protein docking using learned three-dimensional representations. 2019; bioRxiv doi:19 August 2019, preprint: not peer reviewed 10.1101/738690. [DOI]
- 32. Williams C.J., Headd J.J., Moriarty N.W., Prisant M.G., Videau L.L., Deis L.N., Verma V., Keedy D.A., Hintze B.J., Chen V.B.et al.. MolProbity: more and better reference data for improved all-atom structure validation. Protein Sci. 2018; 27:293–315. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The nCoVDock2 web server, with the online documentation, is publicly available at https://ncovdock2.schanglab.org.cn.