Prediction of interactions between the SARS-CoV-2 ORF3a protein and small-molecule ligands using the AND-System cognitive platform, graph neural networks, and molecular

   In recent years, artificial intelligence methods based on the analysis of heterogeneous graphs of biomedical networks have become widely used for predicting molecular interactions. In particular, graph neural networks (GNNs) effectively identify missing edges in gene networks – such as protein–protein interaction, gene–disease, drug–target, and other networks – thereby enabling the prediction of new biological relationships. To reconstruct gene networks, cognitive systems for automatic text mining of scientific publications and databases are often employed. One such AI-driven platform, ANDSystem, is designed for automatic knowledge extraction of molecular interactions and, on this basis, the reconstruction of associative gene networks. The ANDSystem knowledge base contains information on more than 100 million interactions among diverse molecular genetic entities (genes, proteins, metabolites, drugs, etc.). The interactions span a wide range of types: regulatory relationships, physical interactions (protein–protein, protein–ligand), catalytic and chemical reactions, and associations among genes, phenotypes, diseases, and more. In the present study, we applied attention-based graph neural networks trained on the ANDSystem knowledge graph to predict new edges between proteins and ligands and to identify potential ligands for the SARS-CoV-2 ORF3a protein. The accessory protein ORF3a plays an important role in viral pathogenesis through ion-channel activity, induction of apoptosis, and the ability to modulate endolysosomal processes and the host innate immune response. Despite this broad functional spectrum, ORF3a has been explored far less as a pharmacological target than other viral proteins. Using a graph neural network, we predicted five small molecules of different origins (metabolites and a drug) that potentially interact with ORF3a: N-acetyl-D-glucosamine, 4-(benzoylamino)benzoic acid, austocystin D, bictegravirum, and L-threonine. Molecular docking and MM/GBSA affinity estimation indicate the potential ability of these compounds to form complexes with ORF3a. Localization analysis showed that the binding sites of bictegravir and 4-(benzoylamino)benzoic acid lie in a cytosolic surface pocket of the protein that is solvent-exposed; L-threonine binds within the intersubunit cleft of the dimer; and austocystin D and N-acetyl-D-glucosamine are positioned at the boundary between the cytosolic surface and the transmembrane region. The accessibility of these binding sites may be reduced by the influence of the lipid bilayer. The binding energetics for bictegravirum were more favorable than for 4-(benzoylamino)benzoic acid (docking score −7.37 kcal/mol; MM/GBSA ΔG −14.71 ± 3.12 kcal/mol), making bictegravirum a promising candidate for repurposing as an ORF3a inhibitor.

1. Ahsan T., Sajib A.A. Repurposing of approved drugs with potential to interact with SARS-CoV-2 receptor. Biochem Biophys Rep. 2021;26:100982. doi: 10.1016/j.bbrep.2021.100982

2. Ali S.I.M., Alrashid S.Z. A review of methods for gene regulatory networks reconstruction and analysis. Artif Intell Rev. 2025;58:256. doi: 10.1007/s10462­025­11257­z

3. Barberis E., Timo S., Amede E., Vanella V.V., Puricelli C., Cappellano G., Raineri D., … Rolla R., Chiocchetti A., Baldanzi G., Marengo E., Manfredi M. Large-scale plasma analysis revealed new mechanisms and molecules associated with the host response to SARS-CoV-2. Int J Mol Sci. 2020;21(22):8623. doi: 10.3390/ijms21228623

4. Baysal Ö., Abdul Ghafoor N., Silme R.S., Ignatov A.N., Kniazeva V. Molecular dynamics analysis of N-acetyl-D-glucosamine against specific SARS­CoV­2’s pathogenicity factors. PLoS One. 2021; 16(5):e0252571. doi: 10.1371/journal.pone.0252571

5. Boby M.L., Fearon D., Ferla M., Filep M., Koekemoer L., Robinson M.C., COVID Moonshot Consortium, … Zaidmann D., Zhang I., Zidane H., Zitzmann N., Zvornicanin S.N. Open science discovery of potent non-covalent SARS-CoV-2 main protease inhibitors. Science. 2023;380(6640):eabo7201. doi: 10.1126/science.abo7201

6. Bragina E.Y., Tiys E.S., Freidin M.B., Koneva L.A., Demenkov P.S., Ivanisenko V.A., Kolchanov N.A., Puzyrev V.P. Insights into pathophysiology of dystropy through the analysis of gene networks: an example of bronchial asthma and tuberculosis. Immunogenetics. 2014;66(7­8):457­465. doi: 10.1007/s00251­014­0786­1

7. Bragina E.Y., Tiys E.S., Rudko A.A., Ivanisenko V.A., Freidin M.B. Novel tuberculosis susceptibility candidate genes revealed by the reconstruction and analysis of associative networks. Infect Genet Evol. 2016;46:118­123. doi: 10.1016/j.meegid.2016.10.030

8. Butikova E.A., Basov N.V., Rogachev A.D., Gaisler E.V., Ivanisenko V.A., Demenkov P.S., Makarova A.­L.A., … Pokrovsky A.G., Vinokurov N.A., Kanygin V.V., Popik V.M., Shevchenko O.A. Metabolomic and gene networks approaches reveal the role of mitochondrial membrane proteins in response of human melanoma cells to THz radiation. Biochim Biophys Acta Mol Cell Biol Lipids. 2025;1870(2):159595. doi: 10.1016/j.bbalip.2025.159595

9. Case D.A., Aktulga H.M., Belfon K., Cerutti D.S., Andrés Cisneros G., Cruzeiro V.W.D., Forouzesh N., … Roitberg A., Simmerling C.S., York D.M., Nagan M.C., Merz K.M. Jr. AmberTools. J Chem Inf Model. 2023;63(20):6183­6191. doi: 10.1021/acs.jcim.3c01153

10. Chicco D., Jurman G. The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genomics. 2020;21:6. doi: 10.1186/s12864-019-6413­7

11. Clarivate. MetaBase & MetaCore: Early Research Intelligence Solutions. Available at: http://clarivate.com/life-sciences-healthcare/research-development/discovery-development/early-research-intelligence-solutions/

12. Dong M., Galvan Achi J.M., Du R., Rong L., Cui Q. Development of SARS-CoV-2 entry antivirals. Cell Insight. 2024;3(1):100144. doi: 10.1016/j.cellin.2023.100144

13. 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(8):3891­3898. doi: 10.1021/acs.jcim.1c00203

14. Fey M., Lenssen J.E. Fast graph representation learning with PyTorch Geometric. arXiv. 2019. doi: 10.48550/arXiv.1903.02428

15. Gallant J.E., Thompson M., DeJesus E., Voskuhl G.W., Wei X., Zhang H., Martin H. Antiviral activity, safety, and pharmacokinetics of bictegravir as 10-day monotherapy in HIV-1-infected adults. J Acquir Immune Defic Syndr. 2017;75(1):61­66. doi: 10.1097/QAI.0000000000001306

16. Gaudelet T., Day B., Jamasb A.R., Soman J., Regep C., Liu G., Hayter J.B.R., Vickers R., Roberts C., Tang J., Roblin D., Blundell T.L., Bronstein M.M., Taylor-King J.P. Utilizing graph machine learning within drug discovery and development. Brief Bioinform. 2021; 22(6):bbab159. doi: 10.1093/bib/bbab159

17. Glorot X., Bordes A., Bengio Y. Deep sparse rectifier neural networks. In: Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics (AISTATS). 2011;315­323. Available at: https://proceedings.mlr.press/v15/glorot11a/glorot11a.pdf

18. Glotov A.S., Tiys E.S., Vashukova E.S., Pakin V.S., Demenkov P.S., Saik O.V., Ivanisenko T.V., Arzhanova O.N., Mozgovaya E.V., Zainulina M.S., Kolchanov N.A., Baranov V.S., Ivanisenko V.A. Molecular association of pathogenetic contributors to pre-eclampsia (pre-eclampsia associome). BMC Syst Biol. 2015;9(Suppl. 2):S4. doi: 10.1186/1752­0509­9­S2­S4

19. Gwon Y.­D., Strand M., Lindqvist R., Nilsson E., Saleeb M., Elofsson M., Överby A.K., Evander M. Antiviral activity of benzavir-2 against emerging flaviviruses. Viruses. 2020;12(3):351. doi: 10.3390/v12030351

20. Harris J.K. Primer on binary logistic regression. Fam Med Community Health. 2021;9(Suppl. 1):e001290. doi 10.1136/fmch-2021-001290

21. Hinkle J.J., Trychta K.A., Wires E.S., Osborn R.M., Leach J.R., Faraz Z.F., Svarcbahs R., Richie C.T., Dewhurst S., Harvey B.K. Subcellular localization of SARS-CoV-2 E and 3a proteins along the secretory pathway. J Mol Histol. 2025;56(2):98. doi: 10.1007/s10735­025­10375­w

22. Islam M., Strand M., Saleeb M., Svensson R., Baranczewski P., Artursson P., Wadell G., Ahlm C., Elofsson M., Evander M. Anti-Rift Valley fever virus activity in vitro, pre-clinical pharmacokinetics and oral bioavailability of benzavir-2, a broad-acting antiviral compound. Sci Rep. 2018;8:1925. doi: 10.1038/s41598-018-20362-9

23. Ivanisenko T.V., Saik O.V., Demenkov P.S., Ivanisenko N.V., Savostianov A.N., Ivanisenko V.A. AND-Digest: a new web­based module of AND-System for the search of knowledge in the scientific literature. BMC Bioinformatics. 2020;21(Suppl 11):228. doi: 10.1186/s12859-020­03557­8

24. Ivanisenko T.V., Demenkov P.S., Kolchanov N.A., Ivanisenko V.A. The new version of the AND-Digest tool with improved AI-based short names recognition. Int J Mol Sci. 2022;23(23):14934. doi: 10.3390/ijms232314934

25. Ivanisenko T.V., Demenkov P.S., Ivanisenko V.A. An accurate and efficient approach to knowledge extraction from scientific publications using structured ontology models, graph neural networks, and large language models. Int J Mol Sci. 2024;25(21):11811. doi: 10.3390/ijms252111811

26. Ivanisenko V.A., Saik O.V., Ivanisenko N.V., Tiys E.S., Ivanisenko T.V., Demenkov P.S., Kolchanov N.A. ANDSystem: an associative network discovery system for automated literature mining in the field of biology. BMC Syst Biol. 2015;9(Suppl. 2):S2. doi: 10.1186/17520509-9-S2-S2

27. Ivanisenko V.A., Demenkov P.S., Ivanisenko T.V., Mishchenko E.L., Saik O.V. A new version of the AND-System tool for automatic extraction of knowledge from scientific publications with expanded functionality for reconstruction of associative gene networks by considering tissue­specific gene expression. BMC Bioinformatics. 2019; 20(1):34. doi: 10.1186/s12859­018­2567­6

28. Ivanisenko V.A., Gaisler E.V., Basov N.V., Rogachev A.D., Cheresiz S.V., Ivanisenko T.V., Demenkov P.S., … Karpenko T.N., Velichko A.J., Voevoda M.I., Kolchanov N.A., Pokrovsky A.G. Plasma metabolomics and gene regulatory networks analysis reveal the role of nonstructural SARS-CoV-2 viral proteins in metabolic dysregulation in COVID-19 patients. Sci Rep. 2022;12(1):19977. doi: 10.1038/s41598­022­24170­0

29. Ivanisenko V.A., Rogachev A.D., Makarova A.A., Basov N.V., Gaisler E.V., Kuzmicheva I.N., Demenkov P.S., … Kolchanov N.A., Plesko V.V., Moroz G.B., Lomivorotov V.V., Pokrovsky A.G. AI­assisted identification of primary and secondary metabolomic markers for postoperative delirium. Int J Mol Sci. 2024;25(21):11847. doi: 10.3390/ijms252111847

30. Kern D.M., Sorum B., Mali S.S., Hoel C.M., Sridharan S., Remis J.P., Toso D.B., Kotecha A., Bautista D.M., Brohawn S.G. Cryo­EM structure of SARS-CoV-2 ORF3a in lipid nanodiscs. Nat Struct Mol Biol. 2021;28(7):573­582. doi: 10.1038/s41594-021-00619-0

31. Krämer A., Green J., Pollard J. Jr, Tugendreich S. Causal analysis approaches in Ingenuity Pathway Analysis. Bioinformatics. 2014; 30(4):523­530. doi: 10.1093/bioinformatics/btt703

32. Larina I.M., Pastushkova L.Kh., Tiys E.S., Kireev K.S., Kononikhin A.S., Starodubtseva N.L., Popov I.A., Custaud M.-A., Dobrokhotov I.V., Nikolaev E.N., Kolchanov N.A., Ivanisenko V.A. Permanent proteins in the urine of healthy humans during the Mars-500 experiment. J Bioinform Comput Biol. 2015;13(1):1540001. doi: 10.1142/S0219720015400016

33. Lebedeva N.S., Gubarev Y.A., Mamardashvili G.M., Zaitceva S.V., Zdanovich S.A., Malyasova A.S., Romanenko J.V., Koifman M.O., Koifman O.I. Theoretical and experimental study of interaction of macroheterocyclic compounds with ORF3a of SARS-CoV-2. Sci Rep. 2021;11:19481. doi: 10.1038/s41598­021­99072­8

34. Mao A., Mohri M., Zhong Y. Cross­entropy loss functions: theoretical analysis and applications. In: International Conference on Machine Learning (ICML). 2023;23803-23828. Available at: https://proceedings.mlr.press/v202/mao23b/mao23b.pdf

35. Marks K.M., Park E.S., Arefolov A., Russo K., Ishihara K., Ring J.E., Clardy J., Clarke A.S., Pelish E.P. The selectivity of austocystin D arises from cell­line­specific drug activation by cytochrome P450 enzymes. J Nat Prod. 2011;74(4):567­573. doi: 10.1021/np100429s

36. Messina F., Giombini E., Montaldo C., Sharma A.A., Zoccoli A., Sekaly R.P., Locatelli F., Zumla A., Maeurer M., Capobianchi M.R., Lauria F.N., Ippolito G. Looking for pathways related to COVID­19: confirmation of pathogenic mechanisms by SARS­CoV­2­host interactome. Cell Death Dis. 2021;12(8):788. doi: 10.1038/s41419-021-03881-8

37. Miller A.N., Houlihan P.R., Matamala E., Cabezas-Bratesco D., Lee G.Y., Cristofori­Armstrong B., Dilan T.L., Sanchez­Martinez S., Matthies D., Yan R., Yu Z., Ren D., Brauchi S.E., Clapham D.E. The SARS-CoV-2 accessory protein Orf3a is not an ion channel. eLife. 2023;12:e84477. doi: 10.7554/eLife.84477

38. Momynaliev K.T., Kashin S.V., Chelysheva V.V., Selezneva O.V., Demina I.A., Serebryakova M.V., Alexeev D., Ivanisenko V.A., Aman E., Govorun V.M. Functional divergence of Helicobacter pylori related to early gastric cancer. J Proteome Res. 2010;9(1): 254­267. doi: 10.1021/pr900586w

39. Naqvi A.A.T., Fatima K., Muhammad T., Fatima U., Singh I.K., Singh A., Atif S.M., Hariprasad G., Hasan G.M., Hassan M.I. Insights into SARS-CoV-2 genome, structure, evolution, pathogenesis and therapies. Int J Biol Sci. 2020;16(10):1708­1724. doi: 10.7150/ijbs.45127

40. Ng T.I., Correia I., Seagal J., DeGoey D.A., Schrimpf M.R., Hardee D.J., Noey E.L., Kati W.M. Antiviral drug discovery for the treatment of COVID-19 infections. Viruses. 2022;14(5):961. doi: 10.3390/v14050961

41. Nicholson D.N., Greene C.S. Constructing knowledge graphs and their biomedical applications. Comput Struct Biotechnol J. 2020;18: 1414-1421. doi: 10.1016/j.csbj.2020.05.017

42. Oner E., Demirhan I., Miraloglu M., Yalin S., Kurutas E.B. Investigation of antiviral substances in COVID­19 by molecular docking: in silico study. Afr Health Sci. 2023;23(1):23­36. doi: 10.4314/ahs.v23i1.4

43. Páez-Franco J.C., Torres-Ruiz J., Sosa-Hernández V.A., Cervantes-Díaz R., Romero-Ramírez S., Pérez-Fragoso A., Meza-Sánchez D.E., Germán­Acacio J.M., Maravillas­Montero J.L., Mejía­Domínguez N.R., Ponce­de­León A., Ulloa­Aguirre A., Gómez­Martín D., Llorente L. Metabolomics analysis reveals a modified amino acid metabolism that correlates with altered oxygen homeostasis in COVID-19 patients. Sci Rep. 2021;11(1):6350. doi: 10.1038/s41598-021­85788­0

44. Saik O.V., Ivanisenko T.V., Demenkov P.S., Ivanisenko V.A. Interactome of the hepatitis C virus: literature mining with AND-System. Virus Res. 2016;218:40­48. doi: 10.1016/j.virusres.2015.12.003

45. Saik O.V., Nimaev V.V., Usmonov D.B., Demenkov P.S., Ivanisenko T.V., Lavrik I.N., Ivanisenko V.A. Prioritization of genes involved in endothelial cell apoptosis by their implication in lymphedema using an analysis of associative gene networks with AND-System. BMC Med Genomics. 2019;12(Suppl. 2):117. doi: 10.1186/s12920-019-0492-9

46. Sax P.E., Arribas J.R., Orkin C., Lazzarin A., Pozniak A., DeJesus E., Maggiolo F., … Hindman J.T., Martin H., Baeten J.M., Wohl D.; GS­ US­380­1489 and GS­US­380­1490 study investigators. bictegravir/emtricitabine/tenofovir alafenamide as initial treatment for HIV­1: five­year follow-up from two randomized trials. EClinicalMedicine. 2023;59:101991. doi: 10.1016/j.eclinm.2023.101991

47. Stephens E.B., Kunec D., Henke W., Vidal R.M., Greishaber B., Saud R., Kalamvoki M., Singh G., Kafle S., Trujillo J.D., Ferreyra F.M., Morozov I., Richt J.A. The role of the tyrosine-based sorting signals of the ORF3a protein of SARS-CoV-2 in intracellular trafficking and pathogenesis. Viruses. 2025;17(4):522. doi: 10.3390/v17040522

48. Stokes J.M., Yang K., Swanson K., Jin W., Cubillos-Ruiz A., Donghia N.M., MacNair C.R., … Church G.M., Brown E.D., Jaakkola T.S., Barzilay R., Collins J.J. A deep learning approach to antibiotic discovery. Cell. 2020;180(4):688­702. doi: 10.1016/j.cell.2020.01.021

49. Sun C., Zhang J., Wei J., Zheng X., Zhao X., Fang Z., Xu D., Yuan H., Liu Y. Screening, simulation, and optimization design of small-molecule inhibitors of the SARS-CoV-2 spike glycoprotein. PLoS One. 2021;16(1):e0245975. doi: 10.1371/journal.pone.0245975

50. Szklarczyk D., Kirsch R., Koutrouli M., Nastou K., Mehryary F., Hachilif R., Gable A.L., Fang T., Doncheva N.T., Pyysalo S., Bork P., Jensen L.J., von Mering C. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res. 2023; 51(D1):D638­D646. doi: 10.1093/nar/gkac1000

51. Tekin E.D. Investigation of the effects of N­acetylglucosamine on the stability of the spike protein in SARS-CoV-2 by molecular dynamics simulations. Comput Theor Chem. 2023;1222:114049. doi: 10.1016/j.comptc.2023.114049

52. 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(2):455­461. doi: 10.1002/jcc.21334

53. U.S. Food and Drug Administration (FDA). FDA approves first treatment for COVID-­19 (Veklury/remdesivir): press release. 22 Oct 2020. Available at: https://www.fda.gov/news-events/press-announcements/fda-approves-first-treatment-covid-19

54. U.S. Food and Drug Administration. FDA approves first oral antiviral for treatment of COVID-­19 in adults: press announcement. 2023­0525. Available at: https://www.fda.gov/news-events/press-announcements/fda-approves-first-oral-antiviral-treatment-covid-19-adults

55. Veličković P., Cucurull G., Casanova A., Romero A., Liò P., Bengio Y. Graph attention networks. arXiv. 2017. doi: 10.48550/arXiv.1710.10903

56. von Delft A., Hall M.D., Kwong A.D., Purcell L.A., Saikatendu K.S., Schmitz U., Tallarico J.A., Lee A.A. Accelerating antiviral drug discovery: lessons from COVID-­19. Nat Rev Drug Discov. 2023;22(7): 585-603. doi: 10.1038/s41573­023­00692­8

57. Wu Z., Pan S., Chen F., Long G., Zhang C., Yu P.S. A comprehensive survey on graph neural networks. IEEE Trans Neural Netw Learn Syst. 2021;32(1):4­24. doi: 10.1109/TNNLS.2020.2978386

58. Zhang J., Cruz-Cosme R., Zhuang M.W., Liu D., Liu Y., Teng S., Wang P.­H., Tang Q. A systemic and molecular study of subcellular localization of SARS-CoV-2 proteins. Signal Transduct Target Ther. 2021;6(1):192. doi: 10.1038/s41392-021-00564-w

59. Zhang J., Ejikemeuwa A., Gerzanich V., Nasr M., Tang Q., Simard J.M., Zhao R.Y. Understanding the role of SARS-CoV-2 ORF3a in viral pathogenesis and COVID-19. Front Microbiol. 2022;13:854567. doi: 10.3389/fmicb.2022.854567

60. Zhang Y., Sun H., Pei R., Mao B., Zhao Z., Li H., Lin Y., Lu K. The SARS-CoV-2 protein ORF3a inhibits fusion of autophagosomes with lysosomes. Cell Discov. 2021;7:31. doi: 10.1038/s41421-021-00268-z

61. Zhou P., Xie X., Lin Z., Yan S. Towards understanding convergence and generalization of AdamW. IEEE Trans Pattern Anal Mach Intell. 2024;46(9):6486­6493. doi: 10.1109/TPAMI.2024.3382294

62. Zitnik M., Agrawal M., Leskovec J. Modeling polypharmacy side effects with graph convolutional networks. Bioinformatics. 2018; 34(13):i457­i466. doi: 10.1093/bioinformatics/bty294