УДК: 004.8:615.012
DOI: https://doi.org/10.52540/2074-9457.2026.1.72
Скачать статью
А. С. Максимов1, Ж. М. Дергачёва2
ПРИМЕНЕНИЕ МЕТОДОВ МАШИННОГО ОБУЧЕНИЯ В РАННЕЙ РАЗРАБОТКЕ ЛЕКАРСТВЕННЫХ СРЕДСТВ
1ООО «Искамед», г. Минск, Республика Беларусь
2Витебский государственный ордена Дружбы народов медицинский университет, г. Витебск, Республика Беларусь
В статье представлен систематический обзор применения методов машинного обучения на раннем этапе разработки лекарственных средств. Рассмотрены три ключевые стадии: поиск соединений-хитов, установление механизма действия и трансляционные исследования. Показано, что графовые нейронные сети (D-MPNN) и гибридные подходы (Deep Docking) позволяют значительно повысить эффективность виртуального скрининга, а генеративные модели – вариационные автоэнкодеры, нормализующие потоки, диффузионные модели и трансформеры – обеспечивают целенаправленный дизайн новых молекул с заданными свойствами, выходя за пределы существующих химических библиотек. В области предсказания структуры белков проанализирована эволюция от AlphaFold2 и RoseTTAFold к языковым моделям белка (OmegaFold, ESMFold), которые позволяют преодолевать ограничения, связанные с отсутствием эволюционной информации. Рассмотрены диффузионные подходы к молекулярному докингу (DiffDock) и дизайну белков (RFdiffusion), демонстрирующие переход от эвристических методов к непрерывному генеративному моделированию. Отдельное внимание уделено трансферному обучению на примере Geneformer для анализа одноклеточных транскриптомов, а также интеграции прогностических моделей ADMET (ADMET-AI) для многопараметрической оптимизации кандидатов. Сформулированы ключевые вызовы: интерпретируемость моделей, экспериментальная валидация in silico предсказаний, регуляторное признание вычислительных подходов и необходимость развития открытых репозиториев данных высокого качества. Сделан вывод, что методы машинного обучения не заменяют, а дополняют традиционные экспериментальные подходы, формируя основу для нового междисциплинарного направления на стыке химической биологии и data science.
Ключевые слова: машинное обучение, ранняя разработка лекарственных средств, виртуальный скрининг, генеративные модели, графовые нейронные сети, предсказание структуры белков, AlphaFold, языковые модели белка, молекулярный докинг, диффузионные модели, ADMET-прогнозирование, токсичность, трансферное обучение, интерпретируемость моделей.
SUMMARY
A. S. Maksimov, Zh. M. Dergacheva
APPLICATION OF MACHINE LEARNING METHODS IN EARLY DRUG DEVELOPMENT
This article presents a systematic review of machine learning (ML) methods application at an early drug development. Three key phases are considered: top-connections search, mechanism of action elucidation, and translational research. Graph neural networks (D-MPNN) and hybrid approaches (Deep Docking) are shown to substantially enhance virtual screening efficiency, while generative models–variational autoenco-ders normalizing flows, diffusion models and transformers–enable targeted de novo design of molecules with the defined pro-perties leaving beyond existing chemical libraries. In the field of protein structure prediction, the evolution from AlphaFold2 and RoseTTAFold to protein language models (OmegaFold, ESMFold) highlighting their ability to overcome limitations associated with the lack of evolutionary information is analyzed. Diffusion-based approaches to molecular docking (DiffDock) and protein design (RFdiffusion) demonstrating the transition from heuristic methods to continuous generative modeling are discussed. Special attention is given to transfer learning exemplified by Geneformer for single-cell transcriptome analysis, as well as to the integration of ADMET predictive models (ADMET-AI) for multi-parameter optimization of drug candidates. Key challenges are identified: model interpretability, expe-rimental validation of in silico predictions, regulatory acceptance of computational approaches, and the need for high-quality open data repositories. It is concluded that ML methods complement rather than replace traditional experimental approaches, forming the basis for a new interdiscipli-nary field at the intersection of chemical biology and data science.
Keywords: machine learning, early drug development, virtual screening, generative models, graph neural networks, protein structure prediction, AlphaFold, protein language models, molecular docking, diffusion models, ADMET prediction, toxicity, transfer learning, model interpretability.
ЛИТЕРАТУРА
1. Wouters, O. J. Estimated research and development investment needed to bring a new medicine to market, 2009–2018 / O. J. Wouters, M. McKee, J. Luyten // JAMA. – 2020. – Vol. 323, N 9. – P. 844–853. – DOI: 10.1001/jama.2020.1166.
2. Target identification and mechanism of action in chemical biology and drug discovery / M. Schenone, V. Dančik, B. K. Wagner, P. A. Clemons // Nature chemical biology. – 2013. – Vol. 9, N 4. – P. 232–240. – DOI: 10.1038/nchembio.1199.
3. Ashenden, S. K. The Era of Artificial Intelligence, Machine Learning, and Data Science in the Pharmaceutical Industry / S. K. Ashenden. – Cambridge : Academic Press, 2021. – 264 p. – Chap. 6.
4. Smietana, K. Trends in clinical success rates / K. Smietana, M. Siatkowski, M. Møller // Nature reviews. Drug discovery. – 2016. – Vol. 15, N 6. – P. 379–380. – DOI: 10.1038/nrd.2016.85.
5. Harrison, R. K. Phase II and phase III failures: 2013–2015 / R. K. Harrison // Nature reviews. Drug discovery. – 2016. – Vol. 15, N 12. – P. 817–818. – DOI: 10.1038/nrd.2016.184.
6. Dowden, H. Trends in clinical success rates and therapeutic focus / H. Dowden, J. Munro // Nature reviews. Drug discovery. – 2019. – Vol. 18, N 7. – P. 495–496. – DOI: 10.1038/d41573-019-00074-z.
7. Computer vision for autonomous vehicles: problems, datasets and state of the art / J. Janai, F. Güney, A. Behl, A. Geiger // Foundations and Trends in Computer Graphics and Vision. – 2020. – Vol. 12, N 1–3. – P. 1–308. – DOI: 10.1561/0600000079.
8. Machine learning and natural language processing in psychotherapy research: alliance as example use case / S. B. Goldberg, N. Flemotomos, V. R. Martinez [et al.] // Journal of counseling psychology. – 2020. – Vol. 67, N 4. – P. 438–448. – DOI: 10.1037/cou0000382.
9. Peterson, A. A. Small-molecule discovery through DNA-encoded libraries / A. A. Peterson, D. R. Liu // Nature reviews. Drug discovery. – 2023. – Vol. 22, N 9. – P. 699–722. – DOI: 10.1038/s41573-023-00713-6.
10. Machine learning on DNA-encoded library count data using an uncertainty-aware probabilistic loss function / K. S. Lim, A. G. Reidenbach, B. K. Hua [et al.] // Journal of chemical informaition and modeling. – 2022. – Vol. 62, N 10. – P. 2316–2331. – DOI: 10.1021/acs.jcim.2c00041.
11. Machine-learning-based data analysis method for cell-based selection of DNA-encoded libraries / R. Hou, C. Xie, Y. Gui [et al.] // ACS omega. – 2023. – Vol. 8, N 21. – P. 19057–19071. – DOI: 10.1021/acsomega.3c02152.
12. Applications of single-cell RNA sequencing in drug discovery and development / B. Van de Sande, J. S. Lee, E. Mutasa-Gottgens [et al.] // Nature reviews. Drug discovery. – 2023. – Vol. 22, N 6. – P. 496–520. – DOI: 10.1038/s41573-023-00688-4.
13. scBERT as a large-scale pretrained deep language model for cell type annotation of single-cell RNA-seq data / F. Yang, W. Wang, F. Wang [et al.] // Nature machine intelligence. – 2022. – Vol. 4, N 10. – P. 852–866. – DOI: 10.1038/s42256-022-00534-z.
14. Deep transfer learning of cancer drug responses by integrating bulk and single-cell RNA-seq data / J. Chen, X. Wang, A. Ma [et al.] // Nature communications. – 2022. – Vol. 13, N 1. – Art. 6494. – DOI: 10.1038/s41467-022-34277-7.
15. A multi-scale convolutional neural network for phenotyping high-content cellular images / W. J. Godinez, I. Hossain, S. E. Lazic [et al.] // Bioinformatics. – 2017. – Vol. 33, N 13. – P. 2010–2019. – DOI: 10.1093/bioinformatics/btx069.
16. A deep learning approach to antibiotic discovery / J. M. Stokes, K. Yang, K. Swanson [et al.] // Cell. – 2020. – Vol. 180, N 4. – P. 688–702. – DOI: 10.1016/j.cell.2020.01.021.
17. Convolutional embedding of attributed molecular graphs for physical property prediction / C. W. Coley, R. Barzilay, W. H. Green [et al.] // Journal of chemical information and modeling. – 2017. – Vol. 57, N 8. – P. 1757–1772. – DOI: 10.1021/acs.jcim.6b00601.
18. Deep learning identifies synergistic drug combinations for treating COVID-19 / W. Jin, J. M. Stokes, R. T. Eastman [et al.] // Proceedings of the National Academy of Sciences of the United States of America. – 2021. – Vol. 118, N 39. – Art. e2105070118. – DOI: 10.1073/pnas.2105070118.
19. Applications of machine learning in drug discovery and development / J. Vamathevan, D. Clark, P. Czodrowski [et al.] // Nature reviews. Drug discovery. – 2019. – Vol. 18, N 6. – P. 463–477. – DOI: 10.1038/s41573-019-0024-5.
20. Rogers, D. Extended-connectivity fingerprints / D. Rogers, M. Hahn // Journal of chemical information and modeling. – 2010. – Vol. 50, N 5. – P. 742–754. – DOI: 10.1021/ci100050t.
21. Database fingerprint (DFP): an approach to represent molecular databases / E. Fernández-De Gortari, C. R. García-Jacas, K. Martinez-Mayorga, J. L. Medina-Franco // Journal of cheminformatics. – 2017. – Vol. 9, N 1. – Art. 9. – Doi: 10.1186/s13321-017-0195-1.
22. Deep learning-guided discovery of an antibiotic targeting Acinetobacter baumannii / G. Liu, D. B. Catacutan, K. Rathod [et al.] // Nature chemical biology. – 2023. – Vol. 19, N 11. – P. 1342–1350. – DOI: 10.1038/s41589-023-01349-8.
23. Analyzing learned molecular representations for property prediction / K. Yang, K. Swanson, W. Jin [et al.] // Journal of chemical information and modeling. – 2019. – Vol. 59, N 8. – P. 3370–3388. – DOI: 10.1021/acs.jcim.9b00237.
24. The Drug Repurposing Hub: a next-generation drug library and information resource / S. M. Corsello, J. A. Bittker, Z. Liu [et al.] // Nature medicine. – 2017. – Vol. 23, N 4. – P. 405–408. – DOI: 10.1038/nm.4306.
25. Discovery of a structural class of antibiotics with explainable deep learning / F. Wong, E. J. Zheng, J. A. Valeri [et al.] // Nature. – 2024. – Vol. 626, N 7997. – P. 177–185. – DOI: 10.1038/s41586-023-06887-8.
26. A practical guide to large-scale docking / B. J. Bender, S. Gahbauer, A. Luttens [et al.] // Nature protocols. – 2021. – Vol. 16, N 10. – P. 4799–4832. – DOI: 10.1038/s41596-021-00597-z.
27. Artificial intelligence-enabled virtual screening of ultra-large chemical libraries with deep docking / F. Gentile, J. C. Yaacoub, J. Gleave [et al.] // Nature protocols. – 2022. – Vol. 17, N 3. – P. 672–697. – DOI: 10.1038/s41596-021-00659-2.
28. Integrating QSAR modelling and deep learning in drug discovery: the emergence of deep QSAR / A. Tropsha, O. Isayev, A. Varnek [et al.] // Nature reviews. Drug discovery. – 2024. – Vol. 23, N 2. – P. 141–155. – DOI: 10.1038/s41573-023-00832-0.
29. Supercomputer-based ensemble docking drug discovery pipeline with application to Covid-19 / A. Acharya, R. Agarwal, M. B. Baker [et al.] // Journal of chemical information and modeling. – 2020. – Vol. 60, N 12. – P. 5832–5852. – DOI: 10.1021/acs.jcim.0c01010.
30. A critical overview of computational approaches employed for COVID-19 drug discovery / E. N. Muratov, R. Amaro, C. H. Andrade [et al.] // Chemical Society reviews. – 2021. – Vol. 50, N 16. – P. 9121–9151. – DOI: 10.1039/d0cs01065k.
31. Sterling, T. ZINC 15 – ligand discovery for everyone / T. Sterling, J. J. Irwin // Journal of chemical information and modeling. – 2015. – Vol. 55, N 11. – P. 2324–2337. – DOI: 10.1021/acs.jcim.5b00559.
32. Non-covalent SARS-CoV-2 Mpro inhibitors developed from in silico screen hits / G. G. Rossetti, M. A. Ossorio, S. Rempel [et al.] // Scientific reports. – 2022. – Vol. 12, N 1. – Art. 2505. – DOI: 10.1038/s41598-022-06306-4.
33. Reymond, J. L. The chemical space project / J. L. Reymond // Accounts of chemical research. – 2015. – Vol. 48, N 3. – P. 722–730. – DOI: 10.1021/ar500432k.
34. Automatic chemical design using a data-driven continuous representation of molecules / R. Gómez-Bombarelli, J. N. Wei, D. Duvenaud [et al.] // ACS central science. – 2018. – Vol. 4, N 2. – P. 268–276. – DOI: 10.1021/acscentsci.7b00572.
35. Anstine, D. M. Generative models as an emerging paradigm in the chemical sciences / D. M. Anstine, O. Isayev // Journal of the American Chemical Society. – 2023. – Vol. 145, N 16. – P. 8736–8750. – DOI: 10.1021/jacs.2c13467.
36. Jin, W. Junction tree variational autoencoder for molecular graph generation / W. Jin, R. Barzilay, T. Jaakkola // arXiv. – 2018. – 18 Feb. – URL: https://arxiv.org/abs/1802.04364 (date of access: 16.03.2026).
37. Design of potent antimalarials with generative chemistry / W. J. Godinez, E. J. Ma, A. T. Chao [et al.] // Nature machine intelligence. – 2022. – Vol. 4, N 2. – P. 180–186. – DOI: 10.1038/s42256-022-00448-w.
38. Walters, W. P. Assessing the impact of generative AI on medicinal chemistry / W. P. Walters, M. Murcko // Nature biotechnology. – 2020. – Vol. 38, N 2. – P. 143–145. – DOI: 10.1038/s41587-020-0418-2.
39. Deep learning tools to accelerate antibiotic discovery / A. Cesaro, M. Bagheri, M. Torres [et al.] // Expert opinion on drug discovery. – 2023. – Vol. 18, N 11. – P. 1245–1257. – DOI: 10.1080/17460441.2023.2250721.
40. Rezende, D. J. Variational inference with normalizing flows / D. J. Rezende, S. Mohamed // International Conference on Machine Learning : proceedings of the 32nd International Conference on Machine Learning / ed.: F. Bach, D. Blei. – 2015. – Vol. 37. – P. 1530–1538.
41. Shekhovtsov, A. VAE approximation error: ELBO and exponential families / A. Shekhovtsov, D. Schlesinger, B. Flach // arXiv. – 2021. – 18 Feb. – URL: https://arxiv.org/abs/2102.09310 (date of access: 16.03.2026).
42. GraphAF: a flow-based autoregressive model for molecular graph generation / C. Shi, M. Xu, Z. Zhu [et al.] // arXiv. – 2020. – 27 Jan. – URL: https://arxiv.org/abs/2001.09382 (date of access: 16.03.2026).
43. Equivariant diffusion for molecule generation in 3D / E. Hoogeboom, V. G. Satorras, C. Vignac, M. Welling // International Conference on Machine Learning : proceedings of the 39th International Conference on Machine Learning. – 2022. – Vol. 162. – P. 8867–8887.
44. Weininger, D. SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules / D. Weininger // Journal of Chemical Information and Computer Sciences. – 1988. – Vol. 28, N 1. – P. 31–36. – DOI: 10.1021/ci00057a005.
45. Grisoni, F. Chemical language models for de novo drug design: challenges and opportunities / F. Grisoni // Current opinion in structural biology. – 2023. – Vol. 79. – Art. 102527. – DOI: 10.1016/j.sbi.2023.102527.
46. Flam-Shepherd, D. Language models can learn complex molecular distributions / D. Flam-Shepherd, K. Zhu, A. Aspuru-Guzik // Nature communications. – 2022. – Vol. 13, N 1. – Art. 3293. – DOI: 10.1038/s41467-022-30839-x.
47. Chemical language models enable navigation in sparsely populated chemical space / M. A. Skinnider, R. Greg Stacey, D. S. Wishart, L. J. Foster // Nature machine intelligence. – 2021. – Vol. 3, N 9. – P. 759–770. – DOI: 10.1038/s42256-021-00368-1.
48. Generative molecular design in low data regimes / M. Moret, L. Friedrich, F. Grisoni [et al.] // Nature machine intelligence. – 2020. – Vol. 2, N 3 – P. 171–180. – DOI: 10.1038/s42256-020-0160-y.
49. De novo design of Nurr1 agonists via fragment-augmented generative deep learning in low-data regime / M. Ballarotto, S. Willems, T. Stiller [et al.] // Journal of medicinal chemistry. – 2023. – Vol. 66, N 12. – P. 8170–8177. – DOI: 10.1021/acs.jmedchem.3c00485.
50. Leveraging molecular structure and bioactivity with chemical language models for de novo drug design / M. Moret, I. P. Angona, L. Cotos [et al.] // Nature communications. – 2023. – Vol. 14, N 1. – Art. 114. – DOI: 10.1038/s41467-022-35692-6.
51. Combining generative artificial intelligence and on-chip synthesis for de novo drug design / F. Grisoni, B. J. H. Huisman, A. L. Button [et al.] // Science advances. – 2021. – Vol. 7, N 24. – Art. eabg3338. – DOI: 10.1126/sciadv.abg3338.
52. De novo design of bioactive small molecules by artificial intelligence / D. Merk, L. Friedrich, F. Grisoni, G. Schneider // Molecular informatics. – 2018. – Vol. 37, N 1/2. – Art. 1700153. – DOI: 10.1002/minf.201700153.
53. Attention is all you need / A. Vaswani, N. Shazeer, N. Parmar [et al.] // arXiv. – 2017. – 12 June. – URL: https://arxiv.org/abs/1706.03762 (date of access: 16.03.2026).
54. MolGPT: molecular generation using a transformer-decoder model / V. Bagal, R. Aggarwal, P. K. Vinod, U. D. Priyakumar // Journal of chemical information and modeling. – 2021. – Vol. 62, N 9. – P. 2064–2076. – DOI: 10.1021/acs.jcim.1c00600.
55. GuacaMol: benchmarking models for de novo molecular design / N. Brown, M. Fiscato, M. H. S. Segler, A. C. Vaucher // Journal of chemical information and modeling. – 2019. – Vol. 59, N 3. – P. 1096–1108. – DOI: 10.1021/acs.jcim.8b00839.
56. Molecular sets (MOSES): a benchmarking platform for molecular generation models / D. Polykovskiy, A. Zhebrak, B. Sanchez-Lengeling [et al.] // Frontiers in pharmacology. – 2020. – Vol. 11. – Art. 565644. – DOI: 10.3389/fphar.2020.565644.
57. Autonomous chemical research with large language models / D. A. Boiko, R. MacKnight, B. Kline, G. Gomes // Nature. – 2023. – Vol. 624. – P. 570–578. – DOI: 10.1038/s41586-023-06792-0.
58. Leveraging large language models for predictive chemistry / K. M. Jablonka, P. Schwaller, A. Ortega-Guerrero, B. Smit // Nature machine intelligence. – 2024. – Vol. 6, N 2. – P. 161–169. – DOI: 10.1038/s42256-023-00788-1.
59. Born, J. Regression Transformer enables concurrent sequence regression and generation for molecular language modelling / J. Born, M. Manica // Nature machine intelligence. – 2023. – Vol. 5, N 4. – P. 432–444. – DOI: 10.1038/s42256-023-00639-z.
60. Neural scaling of deep chemical models / N. C. Frey, R. Soklaski, S. Axelrod [et al.] // Nature machine intelligence. – 2023. – Vol. 5, N 11. – P. 1297–1305. – DOI: 10.1038/s42256-023-00740-3.
61. Grechishnikova, D. Transformer neural network for protein-specific de novo drug generation as a machine translation problem / D. Grechishnikova // Scientific reports. – 2021. – Vol. 11, N 1. – Art. 321. – DOI: 10.1038/s41598-020-79682-4.
62. Highly accurate protein structure prediction with AlphaFold / J. Jumper, R. Evans, A. Pritzel [et al.] // Nature. – 2021. – Vol. 596, N 7873. – P. 583–589. – DOI: 10.1038/s41586-021-03819-2.
63. Structure of the decoy module of human glycoprotein 2 and uromodulin and its interaction with bacterial adhesin FimH / A. Stsiapanava, C. Xu, S. Nishio [et al.] // Nature structural & molecular biology. – 2022. – Vol. 29, N 3. – P. 190–193. – DOI: 10.1038/s41594-022-00729-3.
64. Cryo-EM structures of human hepatitis B and woodchuck hepatitis virus small spherical subviral particles / H. Liu, X. Hong, J. Xi [et al.] // Science advances. – 2022. – Vol. 8, N 31. – Art. eabo4184. – DOI: 10.1126/sciadv.abo4184.
65. AlphaFold accelerates artificial intelligence powered drug discovery: efficient discovery of a novel CDK20 small molecule inhibitor / F. Ren, X. Ding, M. Zheng [et al.] // Chemical science. – 2023. – Vol. 14, N 6. – P. 1443–1452. – DOI: 10.48550/arXiv.2201.09647.
66. Structural comparison and drug screening of spike proteins of ten SARS-CoV-2 variants / Q. Yang, X. Lian, A. A. S. Syed [et al.] // Research (Washington, D.C.). – 2022. – Vol. 2022. – Art. 9781758. – DOI: 10.34133/2022/9781758.
67. Highly accurate protein structure prediction and drug screen of monkeypox virus proteome / Q. Yang, D. Xia, A. A. S. Syed [et al.] // The Journal of infection. – 2023. – Vol. 86, N 1. – P. 66–117. – DOI: 10.1016/j.jinf.2022.08.006.
68. Chemistry42: an AI-driven platform for molecular design and optimization / Y. A. Ivanenkov, D. Polykovskiy, D. Bezrukov [et al.] // Journal of Chemical Information and Modeling. – 2023. – Vol. 63, N 3. – P. 695–701. – DOI: 10.1021/acs.jcim.2c01191.
69. Accurate prediction of protein structures and interactions using a three-track neural network / M. Baek, F. DiMaio, I. Anishchenko [et al.] // Science. – 2021. – Vol. 373, N 6557. – P. 871–876. – DOI: 10.1126/science.abj8754.
70. The Protein Data Bank / H. M. Berman, J. Westbrook, Z. Feng [et al.] // Nucleic acids research. – 2000. – Vol. 28, N 1. – P. 235–242. – DOI: 10.1093/nar/28.1.235.
71. Van Wart, H. E. The cysteine switch: a principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene family / H. E. Van Wart, H. Birkedal-Hansen // Proceedings of the National Academy of Sciences of the United States of America. – 1990. – Vol. 87, N 14. – P. 5578–5582. – DOI: 10.1073/pnas.87.14.5578.
72. Michaud, J. M. A language model beats AlphaFold2 on orphans / J. M. Michaud, A. Madani, J. S. Fraser // Nature biotechnology. – 2022. – Vol. 40, N 11. – P. 1576–1577. – DOI: 10.1038/s41587-022-01466-0.
73. High-resolution de novo structure prediction from primary sequence / R. Wu, F. Ding, R. Wang [et al.] // bioRxiv. – 2022. – 21 July. – DOI:10.1101/2022.07.21.500999.
74. A method for multiple-sequence-alignment-free protein structure prediction using a protein language model / X. Fang, F. Wang, L. Liu [et al.] // Nature machine intelligence. – 2023. – Vol. 5, N 10. – P. 1087–1096. – DOI: 10.1038/s42256-023-00721-6.
75. Large language models generate functional protein sequences across diverse families / A. Madani, B. Krause, E. R. Greene [et al.] // Nature biotechnology. – 2023. – Vol. 41, N 8. – P. 1099–1106. – DOI: 10.1038/s41587-022-01618-2.
76. Evolutionary-scale prediction of atomic-level protein structure with a language model / Z. Lin, H. Akin, R. Rao [et al.] // Science. – 2023. – Vol. 379, N 6637. – P. 1123–1130. – DOI: 10.1126/science.ade2574.
77. UniRef clusters: a comprehensive and scalable alternative for improving sequence similarity searches / B. E. Suzek, Y. Wang, H. Huang [et al.] // Bioinformatics. – 2015. – Vol. 31, N 6. – P. 926–932. – DOI: 10.1093/bioinformatics/btu739.
78. DiffDock: diffusion steps, twists, and turns for molecular docking / G. Corso, H. Stärk, B. Ling [et al.]. // arXiv. – 2022. – 4 Oct. – URL: https://arxiv.org/abs/2210.01776 (date of access: 16.03.2026).
79. De novo design of protein structure and function with RFdiffusion / J. L. Watson, D. Juergens, N. R. Bennett [et al.] // Nature. – 2023. – Vol. 620, N 7976. – P. 1089–1100. – DOI: 10.1038/s41586-023-06415-8.
80. Transfer learning enables predictions in network biology / C. V. Theodoris, L. Xiao, A. Chopra [et al.] // Nature. – 2023. – Vol. 618, N 7965. – P. 616–624. – DOI: 10.1038/s41586-023-06139-9.
81. Single-nucleus profiling of human dilated and hypertrophic cardiomyopathy / M. Chaffin, I. Papangeli, B. Simonson [et al.] // Nature. – 2022. – Vol. 608, N 7921. – P. 174–180. – DOI: 10.1038/s41586-022-04817-8.
82. Principles of early drug discovery / J. P. Hughes, S. S. Rees, S. B. Kalindjian, K. L. Philpott // British journal of pharmacology. – 2011. – Vol. 162, N 6. – P. 1239–1249. – DOI: 10.1111/j.1476-5381.2010.01127.x.
83. Goodnow, R. A. Hit and lead identification: integrated technology-based approaches / R. A. Goodnow // Drug discovery today. Technologies. – 2006. – Vol. 3, N 4. – P. 367–375. – DOI: 10.1016/j.ddtec.2006.12.009.
84. Transformer-based deep learning method for optimizing ADMET properties of lead compounds / L. Yang, C. Jin, G. Yang [et al.] // Physical chemistry chemical physics. – 2023. – Vol. 25, N 3. – P. 2377–2385. – DOI: 10.1039/d2cp05332b.
85. In silico prediction of hERG blockers using machine learning and deep learning approaches / Y. Chen, X. Yu, W. Li [et al.] // Journal of applied toxicology. – 2023. – Vol. 43, N 10. – P. 1462–1475. – DOI: 10.1002/jat.4477.
86. Accurate clinical toxicity prediction using multi-task deep neural nets and contrastive molecular explanations / B. Sharma, V. Chenthamarakshan, A. Dhurandhar [et al.] // Scientific reports. – 2023. – Vol. 13, N 1. – Art. 4908. – DOI: 10.1038/s41598-023-31169-8.
87. Why 90% of clinical drug development fails and how to improve it? / D. Sun, W. Gao, H. Hu, S. Zhou // Acta Pharm Sin B. – 2022. – Vol. 12, N 7. – P. 3049–3062. – DOI: 10.1016/j.apsb.2022.02.002.
88. Kola, I. Can the pharmaceutical industry reduce attrition rates? / I. Kola, J. Landis // Nature reviews. Drug discovery. – 2004. – Vol. 3, N 8. – P. 711–716. – DOI: 10.1038/nrd1470.
89. Lipinski, C. A. Lead- and drug-like compounds: the rule-of-five revolution / C. A. Lipinski // Drug discovery today. Technologies. – 2004. – Vol. 1, N 4. – P. 337–341. – DOI: 10.1016/j.ddtec.2004.11.007.
90. A robust, viable, and resource sparing HPLC-based log P method applied to common drugs / A. L. Coutinho, R. Gristofoletti, F. Wu [et al.] // International journal of pharmaceutics. – 2023. – Vol. 644. – Art. 123325. – DOI: 10.1016/j.ijpharm.2023.123325.
91. Faller, B. Computational approaches to determine drug solubility / B. Faller, P. Ertl // Advanced drug delivery reviews. – 2007. – Vol. 59, N 7. – P. 533–545. – DOI: 10.1016/j.addr.2007.05.005.
92. Comparison of log P and log D correction models trained with public and proprietary data sets / I. Aliagas, A. Gobbi, M. L. Lee, B. D. Sellers // Journal of computer-aided molecular design. – 2022. – Vol. 36, N 4. – P. 253–262. – DOI: 10.1007/s10822-022-00450-9.
93. Win, Z. M. Using machine learning to predict partition coefficient (log P) and distribution coefficient (log D) with molecular descriptors and liquid chromatography retention time / Z. M. Win, A. M. Y. Cheong, W. S. Hopkins // Journal of chemical information and modeling. – 2023. – Vol. 63, N 7. – P. 1906–1913. – DOI: 10.1021/acs.jcim.2c01373.
94. The METLIN small molecule dataset for machine learning-based retention time prediction / X. Domingo-Almenara, C. Guijas, E. Billings [et al.] // Nature communications. – 2019. – Vol. 10, N 1. – Art. 5811. – DOI: 10.1038/s41467-019-13680-7.
95. ChEMBL: a large-scale bioactivity database for drug discovery / A. Gaulton, L. J. Bellis, A. P. Bento [et al.] // Nucleic acids research. – 2012. – Vol. 40, N Database issue. – P. D1100–D1107. – DOI: 10.1093/nar/gkr777.
96. Datta, R. Efficient lipophilicity prediction of molecules employing deep-learning models / R. Datta, D. Das, S. Das // Chemometrics and intelligent laboratory systems. – 2021. – Vol. 213. – Art. 104309. – DOI: 10.1016/j.chemolab.2021.104309.
97. Prasad, S. A deep learning approach for the blind log P prediction in SAMPL6 challenge / S. Prasad, B. R. Brooks // Journal of computer-aided molecular design. – 2020. – Vol. 34, N 5. – P. 535–542. – DOI: 10.1007/s10822-020-00292-3.
98. Cardiac safety assays / J. Heijman, N. Voigt, L. G. Carlsson, D. Dobrev // Current opinion in pharmacology. – 2014. – Vol. 15. – P. 16–21. – DOI: 10.1016/j.coph.2013.11.004.
99. CACHE (Critical Assessment of Computational Hit-finding Experiments): a public-private partnership benchmarking initiative to enable the development of computational methods for hit-finding / S. Ackloo, R. Al-Awar, R. E. Amaro [et al.] // Nature reviews. Chemistry. – 2022. – Vol. 6, N 4. – P. 287–295. – DOI: 10.1038/s41570-022-00363-z.
100. ADMET-AI: a machine learning ADMET platform for evaluation of large-scale chemical libraries / K. Swanson, P. Walther, J. Leitz [et al.] // Zenodo. – 2023. – 9 Dec. – DOI: 10.5281/zenodo.10372930.
101. MoleculeNet: a benchmark for molecular machine learning / Z. Wu, B. Ramsundar, E. N. Feinberg [et al.] // Chemical science. – 2017. – Vol. 9, N 2. – P. 513–530. – DOI: 10.1039/c7sc02664a.
102. Tox21Challenge to build predictive models of nuclear receptor and stress response pathways as mediated by exposure to environmental chemicals and drugs / R. Huang, M. Xia, D. T. Nguyen [et al.] // Frontiers in environmental science. – 2016. – Vol. 3. – Art. 85. – DOI: 10.3389/fenvs.2015.00085.
103. ZINC-22 – a free multi-billion-scale database of tangible compounds for ligand discovery / B. I. Tingle, K. G. Tang, M. Castanon [et al.] // Journal of chemical information and modeling. – 2023. – Vol. 63, N 4. – P. 1166–1176. – DOI: 10.1021/acs.jcim.2c01253.
104. Quantum chemistry structures and properties of 134 kilo molecules / R. Ramakrishnan, P. O. Dral, M. Rupp, O. A. von Lilienfeld // Scientific data. – 2014. – Vol. 1. – Art. 140022. – DOI: 10.1038/sdata.2014.22.
105. From computer-aided drug discovery to computer-driven drug discovery / L. Frye, S. Bhat, K. Akinsanya, R. Abel // Drug discovery today. Technologies. – 2021. – Vol. 39. – P. 111–117. – DOI: 10.1016/j.ddtec.2021.08.001.
106. High-throughput screening technology in industrial biotechnology / W. Zeng, L. Guo, S. Xu [et al.] // Trends in biotechnology. – 2020. – Vol. 38, N 8. – P. 888–906. – DOI: 10.1016/j.tibtech.2020.01.001.
107. Sarkar, N. Practical applications of machine learning for anti-infective drug discovery / N. Sarkar, J. M. Stokes // Medicinal Chemistry Reviews / ed.: J. Rudolph, J. J. Bronson. – Washington : ACS, 2023. – Vol. 58. – P. 345–375.
108. Applications of machine learning in microbial natural product drug discovery / A. Arnold, J. Alexander, G. Liu,
J. M. Stokes // Expert opinion on drug discovery. – 2023. – Vol. 18, N 11. – P. 1259–1272. – DOI: 10.1080/17460441.2023.2251400.
109. Artificial intelligence for natural product drug discovery / M. W. Mullowney, K. R. Duncan, S. S. Elsayed [et al.] // Nature reviews. Drug discovery. – 2023. – Vol. 22, N 11. – P. 895–916. – DOI: 10.1038/s41573-023-00774-7.
110. Exploiting machine learning for end-to-end drug discovery and development / S. Ekins, A. C. Puhl, K. M. Zorn [et al.] // Nature materials. – 2019. – Vol. 18, N 5. – P. 435–441. – DOI: 10.1038/s41563-019-0338-z.
111. Designing anticancer peptides by constructive machine learning / F. Grisoni, C. S. Neuhaus, G. Gabernet [et al.] // ChemMedChem. – 2018. – Vol. 13, N 13. – P. 1300–1302. – DOI: 10.1002/cmdc.201800204.
112. Chen, J. xDeep-AcPEP: deep learning method for anticancer peptide activity prediction based on convolutional neural network and multitask learning / J. Chen, H. H. Cheong, S. W. I. Siu // Journal of chemical information and modeling. – 2021. – Vol. 61, N 8. – P. 3789–3803. – DOI: 10.1021/acs.jcim.1c00181.
113. Walker, A. S. A machine learning bioinformatics method to predict biological activity from biosynthetic gene clusters / A. S. Walker, J. Clardy // Journal of chemical information and modeling. – 2021. – Vol. 61, N 6. – P. 2560–2571. – DOI: 10.1021/acs.jcim.0c01304.
114. MELLODDY: cross-pharma federated learning at unprecedented scale unlocks benefits in QSAR without compromising proprietary information / W. Heyndrickx, L. Mervin, T. Morawietz [et al.] // Journal of chemical information and modeling. – 2024. – Vol. 64, N 7. – P. 2331–2344. – DOI: 10.1021/acs.jcim.3c00799.
115. A perspective on explanations of molecular prediction models / G. P. Wellawatte, H. A. Gandhi, A. Seshadri, A. D. White // Journal of chemical theory and computation. – 2023. – Vol. 19, N 8. – P. 2149–2160. – DOI: 10.1021/acs.jctc.2c01235.
116. Crowdsourced mapping of unexplored target space of kinase inhibitors / A. Cichońska, B. Ravikumar, R. J. Allaway [et al.] // Nature communications. – 2021. – Vol. 12, N 1. – Art. 3307. – DOI: 10.1101/2019.12.31.891812.
117. Ketkar, N. Deep Learning with Python / N. Ketkar. – Berkeley, CA : Apress, 2017. – P. 97–111.
REFERENCES
1. Wouters OJ, McKee M, Luyten J. Estimated research and development investment needed to bring a new medicine to market, 2009–2018. JAMA. 2020;323(9):844–53. doi: 10.1001/jama.2020.1166
2. Schenone M, Dančik V, Wagner BK, Clemons PA. Target identification and mechanism of action in chemical biology and drug discovery. Nat Chem Biol. 2013;9(4):232–40. doi: 10.1038/nchembio.1199
3. Ashenden SK. The Era of Artificial Intelligence, Machine Learning, and Data Science in the Pharmaceutical Industry, Cambridge, UK: Academic Press; 2021. 264 p. Chapter 6
4. Smietana K, Siatkowski M, Møller M. Trends in clinical success rates. Nat Rev Drug Discov. 2016;15(6):379–80. doi: 10.1038/nrd.2016.85
5. Harrison RK. Phase II and phase III failures: 2013–2015. Nat Rev Drug Discov. 2016;15(12): 817–8. doi: 10.1038/nrd.2016.184
6. Dowden H, Munro J. Trends in clinical success rates and therapeutic focus. Nat Rev Drug Discov. 2019;18(7):495–6. doi: 10.1038/d41573-019-00074-z
7. Janai J, Güney F, Behl A, Geiger A. Computer vision for autonomous vehicles: problems, datasets and state of the art. Foundations and Trends in Computer Graphics and Vision. 2020;12(1–3):1–308. doi: 10.1561/0600000079
8. Goldberg B, Flemotomos N, Martinez VR, Tanana MJ, Kuo PB, Pace BT, et al. Machine learning and natural language processing in psychotherapy research: alliance as example use case. J Couns Psychol. 2020;67(4):438–48. doi: 10.1037/cou0000382
9. Peterson AA, Liu DR. Small-molecule discovery through DNA-encoded libraries. Nat Rev Drug Discov. 2023;22(9):699–722. doi: 10.1038/s41573-023-00713-6
10. Lim KS, Reidenbach AG, Hua BK, Mason JW, Gerry CJ, Clemans PA, et al. Machine learning on DNA-encoded library count data using an uncertainty-aware probabilistic loss function. J Chem Inf Model. 2022;62(10):2316–31. doi: 10.1021/acs.jcim.2c00041
11. Hou R, Xie C, Gui Y, Li G, Li X. Machine-learning-based data analysis method for cell-based selection of DNA-encoded libraries. ACS Omega. 2023:8(21):19057–71. doi: 10.1021/acsomega.3c02152
12. Van de Sande B, Lee JS, Mutasa-Gottgens E, Naughton B, Bacon W, Manning J, et al. Applications of single-cell RNA sequencing in drug discovery and development. Nat Rev Drug Discov. 2023;22(6):496–520. doi: 10.1038/s41573-023-00688-4
13. Yang F, Wang W, Wang F, Fang Y, Tang D, Huang J, et al. scBERT as a large-scale pretrained deep language model for cell type annotation of single-cell RNA-seq data. Nat Mach Intell. 2022;4(10):852–66. doi: 10.1038/s42256-022-00534-z
14. Chen J, Wang X, Ma A, Wang QE, Liu B, Li L, et al. Deep transfer learning of cancer drug responses by integrating bulk and single-cell RNA-seq data. Nat Commun. 2022;13(1 Art 6494). doi: 10.1038/s41467-022-34277-7
15. Godinez WJ, Hossain I, Lazic SE, Davies JW, Zhang X. A multi-scale convolutional neural network for phenotyping high-content cellular images. Bioinformatics. 2017;33(13):2010–9. doi: 10.1093/bioinformatics/btx069
16. Stokes JM, Yang K, Swanson K, Jin W, Cubillos-Ruiz A, Donghia NM, et al. A deep learning approach to antibiotic discovery. Cell. 2020;180(4):688–702. doi: 10.1016/j.cell.2020.01.021
17. Coley CW, Barzilay R, Green WH, Jaakkola TS, Jensen KF. Convolutional embedding of attributed molecular graphs for physical property prediction. J Chem Inf Model. 2017;57(8):1757–72. doi: 10.1021/acs.jcim.6b00601
18. Jin W, Stokes JM, Eastman RT, Itkin Z, Zakharov AV, Collins JJ, et al. Deep learning identifies synergistic drug combinations for treating COVID-19. Proc Natl Acad Sci USA. 2021;118(39 Art e2105070118). doi: 10.1073/pnas.2105070118
19. Vamathevan J, Clark D, Czodrowski P, Dunham I, Ferran E, Lee G, et al. Applications of machine learning in drug discovery and development. Nat Rev Drug Discov. 2019;18(6):463–77. doi: 10.1038/s41573-019-0024-5
20. Rogers D, Hahn M. Extended-connectivity fingerprints. J Chem Inf Model. 2010;50(5):742–54. doi: 10.1021/ci100050t
21. Fernández-De Gortari E, García-Jacas CR, Martinez-Mayorga K, Medina-Franco JL. Database fingerprint (DFP): an approach to represent molecular databases. J Cheminform. 2017;9(1 Art 9). doi: 10.1186/s13321-017-0195-1
22. Liu G, Catacutan DB, Rathod K, Swanson K, Jin W, Mohammed JC, et al. Deep learning-guided discovery of an antibiotic targeting Acinetobacter baumannii. Nat Chem Biol. 2023;19(11):1342–50. doi: 10.1038/s41589-023-01349-8
23. Yang K, Swanson K, Jin W, Coley C, Eiden P, Gao H, et al. Analyzing learned molecular representations for property prediction. J Chem Inf Model. 2019;59(8):3370–88. doi: 10.1021/acs.jcim.9b00237
24. Corsello SM, Bittker JA, Liu Z, Gould J, McCarren P, Hirschman JE, et al. The Drug Repurposing Hub: a next-generation drug library and information resource. Nat Med. 2017;23(4):405–8. doi: 10.1038/nm.4306
25. Wong F, Zheng EJ, Valeri JA, Donghia NM, Anahtar MN, Omori S, et al. Discovery of a structural class of antibiotics with explainable deep learning. Nature. 2024;626(7997):177–85. doi: 10.1038/s41586-023-06887-8
26. Bender BJ, Gahbauer S, Luttens A, Lyu J, Webb CM, Stein RM, et al. A practical guide to large-scale docking. Nat Protoc. 2021;16(10):4799–832. doi: 10.1038/s41596-021-00597-z
27. Gentile F, Yaacoub JC, Gleave J, Fernandez M, Ton AT, Ban F, et al. Artificial intelligence-enabled virtual screening of ultra-large chemical libraries with deep docking. Nat Protoc. 2022;17(3):672–97. doi: 10.1038/s41596-021-00659-2
28. Tropsha A, Isayev O, Varnek A, Schneider G, Cherkasov A. Integrating QSAR modelling and deep learning in drug discovery: the emergence of deep QSAR. Nat Rev Drug Discov. 2024;23(2):141–55. doi: 10.1038/s41573-023-00832-0
29. Acharya A, Agarwal R, Baker MB, Baudry J, Bhowmik D, Boehm S, et al. Supercomputer-based ensemble docking drug discovery pipeline with application to Covid-19. J Chem Inf Model. 2020;60(12):5832–52. doi: 10.1021/acs.jcim.0c01010
30. Muratov EN, Amaro R, Andrade CH, Brown N, Ekins S, Fourches D, et al. A critical overview of computational approaches employed for COVID-19 drug discovery. Chem Soc Rev. 2021;50(16):9121–51. doi: 10.1039/d0cs01065k
31. Sterling T, Irwin JJ. ZINC 15 – ligand discovery for everyone. J Chem Inf Model. 2015;55(11):2324–37. doi: 10.1021/acs.jcim.5b00559
32. Rossetti GG, Ossorio MA, Rempel S, Kratzel A, Dionellis VS, Barriot S, et al. Non-covalent SARS-CoV-2 Mpro inhibitors developed from in silico screen hits. Sci Rep. 2022;12(1 Art 2505. doi: 10.1038/s41598-022-06306-4
33. Reymond JL. The chemical space project. Acc Chem Res. 2015;48(3):722–30. doi: 10.1021/ar500432k
34. Gómez-Bombarelli R, Wei JN, Duvenaud D, Hernandez-Lobato JM, Sanchez-Lengeling B, Sheberla D, et al. Automatic chemical design using a data-driven continuous representation of molecules. ACS Cent Sci. 2018;4(2):268–76. doi: 10.1021/acscentsci.7b00572
35. Anstine DM, Isayev O. Generative models as an emerging paradigm in the chemical sciences. J Am Chem Soc. 2023;145(16):8736–50. doi: 10.1021/jacs.2c13467
36. Jin W, Barzilay R, Jaakkola T. Junction tree variational autoencoder for molecular graph generation. arXiv. 2018 Feb 18. URL: https://arxiv.org/abs/1802.04364 (date of access: 16.03.2026)
37. Godinez WJ, Ma EJ, Chao AT, Pei L, Skewes-Cox P, Canham SM, et al. Design of potent antimalarials with generative chemistry. Nat Mach Intell. 2022;4(2):180–6. doi: 10.1038/s42256-022-00448-w
38. Walters WP, Murcko M. Assessing the impact of generative AI on medicinal chemistry. Nat Biotechnol. 2020;38(2):143–5. doi: 10.1038/s41587-020-0418-2
39. Cesaro A, Bagheri M, Torres M, Wan F, de la Fuente-Nunez C. Deep learning tools to accelerate antibiotic discovery. Expert Opin Drug Discov. 2023;18(11):1245–57. doi: 10.1080/17460441.2023.2250721
40. Rezende DJ, Mohamed S. Variational inference with normalizing flows. In: Bach F, Blei D, editors. International Conference on Machine Learning: proceedings of the 32nd International Conference on Machine Learning. 2015. vol. 37. p. 1530–8
41. Shekhovtsov A, Schlesinger D, Flach B. VAE approximation error: ELBO and exponential families. arXiv. 2021 Feb 18. URL: https://arxiv.org/abs/2102.09310 (date of access: 16.03.2026)
42. Shi C, Xu M, Zhu Z, Zhang W, Zhang M, Tang J. GraphAF: a flow-based autoregressive model for molecular graph generation. arXiv. 2020 Jan 27. URL: https://arxiv.org/abs/2001.09382 (date of access: 16.03.2026)
43. Hoogeboom E, Satorras VG, Vignac C, Welling M. Equivariant diffusion for molecule generation in 3D. In: International Conference on Machine Learning: proceedings of the 39th International Conference on Machine Learning. 2022. vol. 162. p. 8867–87
44. Weininger D. SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules. J Chem Inf Comput Sci. 1988;28(1):31–6. doi: 10.1021/ci00057a005
45. Grisoni F. Chemical language models for de novo drug design: challenges and opportunities. Curr Opin Struct Biol. 2023;79(Art 102527). doi: 10.1016/j.sbi.2023.102527
46. Flam-Shepherd D, Zhu K, Aspuru-Guzik A. Language models can learn complex molecular distributions. Nat Commun. 2022;13(1 Art 3293). doi: 10.1038/s41467-022-30839-x
47. Skinnider MA, Greg Stacey R, Wishart DS, Foster LJ. Chemical language models enable navigation in sparsely populated chemical space. Nat Mach Intell. 2021;3(9):759–70. doi: 10.1038/s42256-021-00368-1
48. Moret M, Friedrich L, Grisoni F, Merk D, Schneider G. Generative molecular design in low data regimes. Nat Mach Intell. 2020;2(3):171–80. doi: 10.1038/s42256-020-0160-y
49. Ballarotto M, Willems S, Stiller T, Nawa F, Marschner JA, Grisoni F, et al. De novo design of Nurr1 agonists via fragment-augmented generative deep learning in low-data regime. J Med Chem. 2023;66(12):8170–7. doi: 10.1021/acs.jmedchem.3c00485
50. Moret M, Angona IP, Cotos L, Yan S, Atz K, Brunner C, et al. Leveraging molecular structure and bioactivity with chemical language models for de novo drug design. Nat Commun. 2023;14(1 Art 114). doi: 10.1038/s41467-022-35692-6
51. Grisoni F, Huisman BJH, Button AL, Moret M, Atz K, Merk D, et al. Combining generative artificial intelligence and on-chip synthesis for de novo drug design. Sci Adv. 2021;7(24 Art eabg3338). doi: 10.1126/sciadv.abg3338
52. Merk D, Friedrich L, Grisoni F, Schneider G. De novo design of bioactive small molecules by artificial intelligence. Mol Inform. 2018;37(1-2 Art 1700153). doi: 10.1002/minf.201700153
53. Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, et al. Attention is all you need. arXiv. 2017 June 12. URL: https://arxiv.org/abs/1706.03762 (date of access: 16.03.2026)
54. Bagal V, Aggarwal R, Vinod PK, Priyakumar UD. MolGPT: molecular generation using a transformer-decoder model. J Chem Inf Model. 2021;62(9):2064–76. doi: 10.1021/acs.jcim.1c00600
55. Brown N, Fiscato M, Segler MHS, Vaucher AC. GuacaMol: benchmarking models for de novo molecular design. J Chem Inf Model. 2019;59(3):1096–108. doi: 10.1021/acs.jcim.8b00839
56. Polykovskiy D, Zhebrak A, Sanchez-Lengeling B, Golovanov S, Tatanov O, Belyaev S, et al. Molecular sets (MOSES): a benchmarking platform for molecular generation models. Front Pharmacol. 2020;11(Art 565644). doi: 10.3389/fphar.2020.565644
57. Boiko DA, MacKnight R, Kline B, Gomes G. Autonomous chemical research with large language models. Nature. 2023;624:570–8. doi: 10.1038/s41586-023-06792-0
58. Jablonka KM, Schwaller P, Ortega-Guerrero A, Smit B. Leveraging large language models for predictive chemistry. Nat Mach Intell. 2024;6(2):161–9. doi: 10.1038/s42256-023-00788-1
59. Born J, Manica M. Regression Transformer enables concurrent sequence regression and generation for molecular language modeling. Nat Mach Intell. 2023;5(4):432–44. doi: 10.1038/s42256-023-00639-z
60. Frey NC, Soklaski R, Axelrod S, Samsi S, G’omez-Bombarelli R, Coley CW, et al. Neural scaling of deep chemical models. Nat Mach Intell. 2023;5(11):1297–305. doi: 10.1038/s42256-023-00740-3
61. Grechishnikova D. Transformer neural network for protein-specific de novo drug generation as a machine translation problem. Sci Rep. 2021;11(1 Art 321). doi: 10.1038/s41598-020-79682-4
62. Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583–9. doi: 10.1038/s41586-021-03819-2
63. Stsiapanava A, Xu C, Nishio S, Han L, Yamakawa N, Carroni M, et al. Structure of the decoy module of human glycoprotein 2 and uromodulin and its interaction with bacterial adhesin FimH. Nat Struct Mol Biol. 2022;29(3):190–3. doi: 10.1038/s41594-022-00729-3
64. Liu H, Hong X, Xi J, Menne S, Hu J, Wang JC. Cryo-EM structures of human hepatitis B and woodchuck hepatitis virus small spherical subviral particles. Sci Adv. 2022;8(31 Art eabo4184). doi: 10.1126/sciadv.abo4184
65. Ren F, Ding X, Zheng M, Korzinkin M, Cai X, Zhu W, et al. AlphaFold accelerates artificial intelligence powered drug discovery: efficient discovery of a novel CDK20 small molecule inhibitor. Chem Sci. 2023;14(6):1443–52. doi: 10.48550/arXiv.2201.09647
66. Yang Q, Lian X, Syed AAS, Fahira A, Zheng C, Zhu Z, et al. Structural comparison and drug screening of spike proteins of ten SARS-CoV-2 variants. Research (Wash D C). 2022;2022(Art 9781758). doi: 10.34133/2022/9781758
67. Yang Q, Xia D, Syed AAS, Wang Z, Shi Y. Highly accurate protein structure prediction and drug screen of monkeypox virus proteome. J Infect. 2023;86(1):66–117. doi: 10.1016/j.jinf.2022.08.006
68. Ivanenkov YA, Polykovskiy D, Bezrukov D, Zagribelnyy B, Aladinskiy V, Kamya P, et al. Chemistry42: an AI-driven platform for molecular design and optimization. J Chem Inf Model. 2023;63(3):695–701. doi: 10.1021/acs.jcim.2c01191
69. Baek M, DiMaio F, Anishchenko I, Dauparas J, Ovchinnikov S, Lee GR, et al. Accurate prediction of protein structures and interactions using a three-track neural network. Science. 2021;373(6557):871–6. doi: 10.1126/science.abj8754
70. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TM, Weissig H, et al. The Protein Data Bank. Nucleic Acids Res. 2000;28(1):235–42. doi: 10.1093/nar/28.1.235
71. Van Wart HE, Birkedal-Hansen H. The cysteine switch: a principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene family. Proc Natl Acad Sci U S A. 1990;87(14):5578–82. doi: 10.1073/pnas.87.14.5578
72. Michaud JM, Madani A, Fraser JS. A language model beats AlphaFold2 on orphans. Nat Biotechnol. 2022;40(11):1576–7. doi: 10.1038/s41587-022-01466-0
73. Wu R, Ding F, Wang R, Shen R, Zhang X, Luo S, et al. High-resolution de novo structure prediction from primary sequence. bioRxiv. 2022 July 21. doi:10.1101/2022.07.21.500999
74. Fang X, Wang F, Liu L, He J, Lin D, Xiang Y, et al. A method for multiple-sequence-alignment-free protein structure prediction using a protein language model. Nat Mach Intell. 2023;5(10):1087–96. doi: 10.1038/s42256-023-00721-6
75. Madani A, Krause B, Greene ER, Subramanian S, Mohr BP, Holton JM, et al. Large language models generate functional protein sequences across diverse families. Nat Biotechnol. 2023;41(8):1099–106. doi: 10.1038/s41587-022-01618-2
76. Lin Z, Akin H, Rao R, Hie B, Zhu Z, Lu W, et al. Evolutionary-scale prediction of atomic-level protein structure with a language model. Science. 2023;379(6637):1123–30. doi: 10.1126/science.ade2574
77. Suzek BE, Wang Y, Huang H, McGarvey PB, Wu CH. UniRef clusters: a comprehensive and scalable alternative for improving sequence similarity searches. Bioinformatics. 2015;31(6):926–32. doi: 10.1093/bioinformatics/btu739
78. Corso G, Stärk H, Ling B, Barzilay R, Jaakkola T. DiffDock: diffusion steps, twists, and turns for molecular docking. arXiv. 2022 Oct 4. URL: https://arxiv.org/abs/2210.01776 (date of access: 16.03.2026)
79. Watson JL, Juergens D, Bennett NR, Trippe BL, Yit J, Eisenach HE, et al. De novo design of protein structure and function with RFdiffusion. Nature;2023;620(7976):1089–100. doi: 10.1038/s41586-023-06415-8
80. Theodoris CV, Xiao L, Chopra A, Chaffin MD, Al Sayed ZR, Hill MC, et al. Transfer learning enables predictions in network biology. Nature. 2023;618(7965):616–24. doi: 10.1038/s41586-023-06139-9
81. Chaffin M, Papangeli I, Simonson B, Akkad AD, Hill MC, Arduini A, et al. Singlenucleus profiling of human dilated and hypertrophic cardiomyopathy. Nature. 2022;608(7921):174–80. doi: 10.1038/s41586-022-04817-8
82. Hughes JP, Rees SS, Kalindjian SB, Philpott KL. Principles of early drug discovery. Br J Pharmacol. 2011;162(6):1239–49. doi: 10.1111/j.1476-5381.2010.01127.x
83. Goodnow RA. Hit and lead identification: integrated technology-based approaches. Drug Discov Today Technol. 2006;3(4):367–75. doi: 10.1016/j.ddtec.2006.12.009
84. Yang L, Jin C, Yang G, Bing Z, Huang L, Niu Y, et al. Transformer-based deep learning method for optimizing ADMET properties of lead compounds. Phys Chem Chem Phys. 2023;25(3):2377–85. doi: 10.1039/d2cp05332b
85. Chen Y, Yu X, Li W, Tang Y, Liu G. In silico prediction of hERG blockers using machine learning and deep learning approaches. J Appl Toxicol. 2023;43(10):1462–75. doi: 10.1002/jat.4477
86. Sharma B, Chenthamarakshan V, Dhurandhar A, Pereira S, Hendler JA, Dordick JS, et al. Accurate clinical toxicity prediction using multi-task deep neural nets and contrastive molecular explanations. Sci Rep. 2023;13(1 Art 4908). doi: 10.1038/s41598-023-31169-8
87. Sun D, Gao W, Hu H, Zhou S. Why 90% of clinical drug development fails and how to improve it? Acta Pharm Sin B. 2022;12(7):3049–62. doi: 10.1016/j.apsb.2022.02.002
88. Kola I, Landis J. Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov. 2004;3(8):711–6. doi: 10.1038/nrd1470
89. Lipinski C. A. Lead- and drug-like compounds: the rule-of-five revolution. Drug Discov Today Technol. 2004;1(4):337–41. doi: 10.1016/j.ddtec.2004.11.007
90. Coutinho AL, Gristofoletti R, Wu F, Shoyaib AA, Dressman J, Polli JE. A robust, viable, and resource sparing HPLC-based log P method applied to common drugs. Int J Pharm. 2023;644(Art 123325). doi: 10.1016/j.ijpharm.2023.123325
91. Faller B, Ertl P. Computational approaches to determine drug solubility. Adv Drug Deliv Rev. 2007;59(7):533–45. doi: 10.1016/j.addr.2007.05.005
92. Aliagas I, Gobbi A, Lee ML, Sellers BD. Comparison of log P and log D correction models trained with public and proprietary data sets. J Comput Aided Mol Des. 2022;36(4):253–62. doi: 10.1007/s10822-022-00450-9
93. Win ZM, Cheong AMY, Hopkins WS. Using machine learning to predict partition coefficient (log P) and distribution coefficient (log D) with molecular descriptors and liquid chromatography retention time. J Chem Inf Model. 2023;63(7):1906–13. doi: 10.1021/acs.jcim.2c01373
94. Domingo-Almenara X, Guijas C, Billings E, Montenegro-Burke JR, Uritboonthai W, Aisporna AE, et al. The METLIN small molecule dataset for machine learning-based retention time prediction. Nat Commun. 2019;10(1 Art 5811). doi: 10.1038/s41467-019-13680-7
95. Gaulton A, Bellis LJ, Bento AP, Gambers J, Davies M, Hersey A, et al. ChEMBL: a large-scale bioactivity database for drug discovery. Nucleic Acids Res. 2012;40(Database issue):D1100–7. doi: 10.1093/nar/gkr777
96. Datta R, Das D, Das S. Efficient lipophilicity prediction of molecules employing deep-learning models. Chemometr Intell Lab Syst. 2021;213(Art 104309). doi: 10.1016/j.chemolab.2021.104309
97. Prasad S, Brooks BR. A deep learning approach for the blind log P prediction in SAMPL6 challenge. J Comput Aided Mol Des. 2020;34(5):535–42. doi: 10.1007/s10822-020-00292-3
98. Heijman J, Voigt N, Carlsson LG, Dobrev D. Cardiac safety assays. Curr Opin Pharmacol. 2014;15:16–21. doi: 10.1016/j.coph.2013.11.004
99. Ackloo S, Al-Awar R, Amaro RE, Arrowsmith CH, Azevedo H, Batey RA, et al. CACHE (Critical Assessment of Computational Hit-finding Experiments): a public-private partnership benchmarking initiative to enable the development of computational methods for hit-finding. Nat Rev Chem. 2022;6(4):287–95. doi: 10.1038/s41570-022-00363-z
100. Swanson K, Walther P, Leitz J, Mukherjee S, Wu JC, Shivnaraine RV, et al. ADMET-AI: a machine learning ADMET platform for evaluation of large-scale chemical libraries. Zenodo. 2023 Dec 9. doi: 10.5281/zenodo.10372930
101. Wu Z, Ramsundar B, Feinberg EN, Gomes J, Geniesse C, Pappu AS, et al. MoleculeNet: a benchmark for molecular machine learning. Chem Sci. 2017;9(2):513–30. doi: 10.1039/c7sc02664a
102. Huang R, Xia M, Nguyen DT, Zhao T, Sakamuru S, Zhao J, et al. Tox21Challenge to build predictive models of nuclear receptor and stress response pathways as mediated by exposure to environmental chemicals and drugs. Front Environ Sci. 2016;3(Art 85). doi: 10.3389/fenvs.2015.00085
103. Tingle BI, Tang KG, Castanon M, Gutierrez JJ, Khurelbaatar M, Dandarchuluun C, et al. ZINC-22 – a free multi-billion-scale database of tangible compounds for ligand discovery. J Chem Inf Model. 2023;63(4):1166–76. doi: 10.1021/acs.jcim.2c01253
104. Ramakrishnan R, Dral PO, Rupp M, von Lilienfeld OA. Quantum chemistry structures and properties of 134 kilo molecules. Sci Data. 2014;1(Art 140022). doi: 10.1038/sdata.2014.22
105. Frye L, Bhat S, Akinsanya K, Abel R. From computer-aided drug discovery to computer-driven drug discovery. Drug Discov Today Technol. 2021;39:111–7. doi: 10.1016/j.ddtec.2021.08.001
106. Zeng W, Guo L, Xu S, Chen J, Zhou J. High-throughput screening technology in industrial biotechnology. Trends Biotechnol. 2020;38(8):888–906. doi: 10.1016/j.tibtech.2020.01.001
107. Sarkar N, Stokes JM. Practical applications of machine learning for anti-infective drug discovery. In: Rudolph J, Bronson JJ, editors. Medicinal Chemistry Reviews. Washington, USA: ACS; 2023. vol. 58. p. 345–75
108. Arnold A, Alexander J, Liu G, Stokes JM. Applications of machine learning in microbial natural product drug discovery. Expert Opin Drug Discov. 2023;18(11):1259–72. doi: 10.1080/17460441.2023.2251400
109. Mullowney MW, Duncan KR, Elsayed SS, Garg N, van der Hooft JJJ, Martin NI, et al. Artificial intelligence for natural product drug discovery. Nat Rev Drug Discov. 2023;22(11):895–916. doi: 10.1038/s41573-023-00774-7
110. Ekins S, Puhl AC, Zorn KM, Lane TR, Russo DP, Klein JJ, et al. Exploiting machine learning for end-to-end drug discovery and development. Nat Mater. 2019;18(5):435–41. doi: 10.1038/s41563-019-0338-z
111. Grisoni F, Neuhaus CS, Gabernet G, Muller AT, Hiss JA, Schneider G. Designing anticancer peptides by constructive machine learning. ChemMedChem. 2018;13(13):1300–2. doi: 10.1002/cmdc.201800204
112. Chen J, Cheong HH, S. Siu SWI. xDeep-AcPEP: deep learning method for anticancer peptide activity prediction based on convolutional neural network and multitask learning. J Chem Inf Model. 2021;61(8):3789–803. doi: 10.1021/acs.jcim.1c00181
113. Walker AS, Clardy J. A machine learning bioinformatics method to predict biological activity from biosynthetic gene clusters. J Chem Inf Model. 2021;61(6):2560–71. doi: 10.1021/acs.jcim.0c01304
114. Heyndrickx W, Mervin L, Morawietz T, Sturm N, Friedrich L, Zalewski A, et al. MELLODDY: cross-pharma federated learning at unprecedented scale unlocks benefits in QSAR without compromising proprietary information. J Chem Inf Model. 2024;64(7):2331–44. doi: 10.1021/acs.jcim.3c00799
115. Wellawatte GP, Gandhi HA, Seshadri A, White AD. A perspective on explanations of molecular prediction models. J Chem Theory Comput. 2023;19(8):2149–60. doi: 10.1021/acs.jctc.2c01235
116. Cichońska A, Ravikumar B, Allaway RJ, Wan F, Park S, Isayev O, et al. Crowdsourced mapping of unexplored target space of kinase inhibitors. Nat Commun. 2021;12(1 Art 3307). doi: 10.1101/2019.12.31.891812
117. Ketkar N. Deep Learning with Python. Berkeley, CA: Apress; 2017. p. 97–111
Адрес для корреспонденции:
210009, Республика Беларусь,
г. Витебск, пр. Фрунзе, 27,
УО «Витебский государственный ордена
Дружбы народов медицинский университет»,
кафедра фармацевтической и
токсикологической химии,
тем. моб.: +375 29 711 50 83,
Дергачёва Ж. М.
Поступила 23.03.2026 г.