Skip to content
2000
Volume 21, Issue 3
  • ISSN: 1573-4099
  • E-ISSN: 1875-6697

Abstract

Traditional molecular generation methods, such as evolutionary algorithms, generate new molecules mainly by linking existing atomic building blocks. The challenging issues in these methods include difficulty in synthesis, failure to achieve desired properties, and structural optimization requirements. Advances in deep learning offer new ideas for rational and robust drug design. Deep learning, a branch of machine learning, is more efficient than traditional methods for processing problems, such as speech, image, and translation. This study provides a comprehensive overview of the current state of research in drug design based on deep learning and identifies key areas for further development. Deep learning-based drug design is pivotal in four key dimensions. Molecular databases form the basis for model training, while effective molecular representations impact model performance. Common DL models (GANs, RNNs, VAEs, CNNs, DMs) generate drug molecules with desired properties. The evaluation metrics guide research directions by determining the quality and applicability of generated molecules. This abstract highlights the foundational aspects of DL-based drug design, offering a concise overview of its multifaceted contributions. Consequently, deep learning in molecule generation has attracted more attention from academics and industry. As a result, many deep learning-based molecule generation types have been actively proposed.

© 2025 Bentham Science Publishers
Loading

Article metrics loading...

/content/journals/cad/10.2174/0115734099287389240126072433
2024-02-06
2025-09-28

  • From This Site

    /content/journals/cad/10.2174/0115734099287389240126072433
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. DiMasiJ.A. GrabowskiH.G. HansenR.W. Innovation in the pharmaceutical industry: New estimates of R&D costs.J. Health Econ.201647203310.1016/j.jhealeco.2016.01.012 26928437
     [Google Scholar]
  2. ErtlP. SchuffenhauerA. Estimation of synthetic accessibility score of drug-like molecules based on molecular complexity and fragment contributions.J. Cheminform.200911810.1186/1758‑2946‑1‑8 20298526
     [Google Scholar]
  3. AguilarR. SpencerK.B. KesnerB. RizviN.F. BadmaliaM.D. MrozowichT. MortisonJ.D. RiveraC. SmithG.F. BurchardJ. DandlikerP.J. PatelT.R. NickbargE.B. LeeJ.T. Targeting Xist with compounds that disrupt RNA structure and X inactivation.Nature2022604790416016610.1038/s41586‑022‑04537‑z 35355011
     [Google Scholar]
  4. SchneiderG. Automating drug discovery.Nat. Rev. Drug Discov.20181729711310.1038/nrd.2017.232 29242609
     [Google Scholar]
  5. BlumL.C. ReymondJ.L. 970 million druglike small molecules for virtual screening in the chemical universe database GDB-13.J. Am. Chem. Soc.2009131258732873310.1021/ja902302h 19505099
     [Google Scholar]
  6. PolishchukP.G. MadzhidovT.I. VarnekA. Estimation of the size of drug-like chemical space based on GDB-17 data.J. Comput. Aided Mol. Des.201327867567910.1007/s10822‑013‑9672‑4 23963658
     [Google Scholar]
  7. MouchlisV.D. MelagrakiG. ZachariaL.C. AfantitisA. Computer-aided drug design of β-secretase, γ-secretase and anti-tau inhibitors for the discovery of novel alzheimer’s therapeutics.Int. J. Mol. Sci.202021370310.3390/ijms21030703 31973122
     [Google Scholar]
  8. SchneiderG. ClarkD.E. Automated de novo drug design: Are we nearly there yet?Angew. Chem. Int. Ed.20195832107921080310.1002/anie.201814681 30730601
     [Google Scholar]
  9. Hassan BaigM. AhmadK. RoyS. Mohammad AshrafJ. AdilM. Haris SiddiquiM. KhanS. Amjad KamalM. ProvazníkI. ChoiI. Computer aided drug design: Success and limitations.Curr. Pharm. Des.201622557258110.2174/1381612822666151125000550 26601966
     [Google Scholar]
  10. JozefowiczR. VinyalsO. SchusterM. ShazeerN. WuY. Exploring the limits of language modeling.arXiv2016201602410
     [Google Scholar]
  11. GravesA. EckD. BeringerN. SchmidhuberJ. Biologically Plausible Speech Recognition with LSTM Neural Nets.Springer200410.1007/978‑3‑540‑27835‑1_10
     [Google Scholar]
  12. KrizhevskyA. SutskeverI. HintonG.E. ImageNet classification with deep convolutional neural networks.Commun. ACM20176068410.1145/3065386
     [Google Scholar]
  13. Gómez-BombarelliR. WeiJ.N. DuvenaudD. Hernández-LobatoJ.M. Sánchez-LengelingB. SheberlaD. Aguilera-IparraguirreJ. HirzelT.D. AdamsR.P. Aspuru-GuzikA. Automatic chemical design using a data-driven continuous representation of molecules.ACS Cent. Sci.20184226827610.1021/acscentsci.7b00572 29532027
     [Google Scholar]
  14. LiY. ZhangL. LiuZ. Multi-objective de novo drug design with conditional graph generative model.J. Cheminform.20181013310.1186/s13321‑018‑0287‑6 30043127
     [Google Scholar]
  15. YangX. ZhangJ. YoshizoeK. TerayamaK. TsudaK. ChemT.S. An efficient python library for de novo molecular generation.Sci. Technol. Adv. Mater.201718197297610.1080/14686996.2017.1401424 29435094
     [Google Scholar]
  16. MercadoR. RastemoT. LindelöfE. KlambauerG. EngkvistO. ChenH. Jannik BjerrumE. Graph networks for molecular design.Mach. Learn. Sci. Technol.20212202502310.1088/2632‑2153/abcf91
     [Google Scholar]
  17. WangM. WangZ. SunH. WangJ. ShenC. WengG. ChaiX. LiH. CaoD. HouT. Deep learning approaches for de novo drug design: An overview.Curr. Opin. Struct. Biol.20227213514410.1016/j.sbi.2021.10.001 34823138
     [Google Scholar]
  18. HoogeboomE. SatorrasV.G. VignacC. WellingM. Equivariant diffusion for molecule generation in 3d. International conference on machine learning. BaltimoreMaryland, USA 202288678887
     [Google Scholar]
  19. SchneuingA. DuY. HarrisC. JamasbA. IgashovI. DuW. BlundellT. LióP. GomesC. WellingM. Structurebased drug design with equivariant diffusion models.arXiv2022202213695
     [Google Scholar]
  20. KusnerM.J. PaigeB. Hern A Ndez-LobatoJ.E.M. Grammar Variational Autoencoder.ICML2017
     [Google Scholar]
  21. LimJ. RyuS. KimJ.W. KimW.Y. Molecular generative model based on conditional variational autoencoder for de novo molecular design.J. Cheminform.20181013110.1186/s13321‑018‑0286‑7 29995272
     [Google Scholar]
  22. LiuQ. AllamanisM. BrockschmidtM. GauntA.L. Constrained graph variational autoencoders for molecule design.Proceedings of the 32nd International Conference on Neural Information Processing SystemsMontréal, Canada2018
     [Google Scholar]
  23. JinW. BarzilayR. JaakkolaT. Junction tree variational autoencoder for molecular graph generation.Proceedings of the 35 th International Conference on Machine LearningSweden2018
     [Google Scholar]
  24. SamantaB. DeA. JanaG. GómezV. ChattarajP. GangulyN. Gomez-RodriguezM. Nevae: A deep generative model for molecular graphs.J. Mach. Learn. Res.20202111413310.5555/3455716.3455830
     [Google Scholar]
  25. SkalicM. SabbadinD. SattarovB. SciabolaS. De FabritiisG. From target to drug: Generative modeling for the multimodal structure-based ligand design.Mol. Pharm.201916104282429110.1021/acs.molpharmaceut.9b00634
     [Google Scholar]
  26. SattarovB. BaskinI.I. HorvathD. MarcouG. BjerrumE.J. VarnekA. De novo molecular design by combining deep autoencoder recurrent neural networks with generative topographic mapping.J. Chem. Inf. Model.20195931182119610.1021/acs.jcim.8b00751 30785751
     [Google Scholar]
  27. ZhavoronkovA. IvanenkovY.A. AliperA. VeselovM.S. AladinskiyV.A. AladinskayaA.V. TerentievV.A. PolykovskiyD.A. KuznetsovM.D. AsadulaevA. Deep learning enables rapid identification of potent DDR1 kinase inhibitors.Nat. Biotechnol.20193791038104010.1038/s41587‑019‑0224‑x
     [Google Scholar]
  28. van DeursenR. ErtlP. TetkoI.V. GodinG. GEN: Highly efficient SMILES explorer using autodidactic generative examination networks.J. Cheminform.20201212210.1186/s13321‑020‑00425‑8 33430998
     [Google Scholar]
  29. MoretM. FriedrichL. GrisoniF. MerkD. SchneiderG. Generative molecular design in low data regimes.Nat. Mach. Intell.20202317118010.1038/s42256‑020‑0160‑y
     [Google Scholar]
  30. KotsiasP.C. Arús-PousJ. ChenH. EngkvistO. TyrchanC. BjerrumE.J. Direct steering of de novo molecular generation with descriptor conditional recurrent neural networks.Nat. Mach. Intell.20202525426510.1038/s42256‑020‑0174‑5
     [Google Scholar]
  31. LvQ. ChenG. ZhaoL. ZhongW. Yu-ChianChen C. Mol2Context-vec: Learning molecular representation from context awareness for drug discovery.Brief. Bioinform.2021226bbab31710.1093/bib/bbab317 34428290
     [Google Scholar]
  32. GebauerN.W.A. GasteggerM. HessmannS.S.P. MüllerK.R. SchüttK.T. Inverse design of 3d molecular structures with conditional generative neural networks.Nat. Commun.202213197310.1038/s41467‑022‑28526‑y 35190542
     [Google Scholar]
  33. OlivecronaM. BlaschkeT. EngkvistO. ChenH. Molecular de-novo design through deep reinforcement learning.J. Cheminform.2017914810.1186/s13321‑017‑0235‑x
     [Google Scholar]
  34. BlaschkeT. Arús-PousJ. ChenH. MargreitterC. TyrchanC. EngkvistO. PapadopoulosK. PatronovA. REINVENT 2.0: An AI tool for de novo drug design.J. Chem. Inf. Model.202060125918592210.1021/acs.jcim.0c00915 33118816
     [Google Scholar]
  35. LoefflerH. HeJ. TiboA. JanetJ.P. VoronovA. MervinL. EngkvistO. REINVENT4: Modern AI-driven generative molecule design.ChemRxiv20232023xt65x10.26434/chemrxiv‑2023‑xt65x
     [Google Scholar]
  36. PopovaM. IsayevO. TropshaA. Deep reinforcement learning for de novo drug design.Sci. Adv.201847eaap788510.1126/sciadv.aap7885 30050984
     [Google Scholar]
  37. LiuX. YeK. van VlijmenH. IJzermanA.P. van WestenG. An exploration strategy improves the diversity of de novo ligands using deep reinforcement learning: A case for the adenosine A(2A) receptor.J. Cheminform.20191113510.1186/s13321‑019‑0355‑6
     [Google Scholar]
  38. LiuX. YeK. van VlijmenH.W.T. EmmerichM.T.M. IJzermanA.P. van WestenG.J.P. DrugEx v2: De novo design of drug molecules by Pareto-based multi-objective reinforcement learning in polypharmacology.J. Cheminform.20211318510.1186/s13321‑021‑00561‑9 34772471
     [Google Scholar]
  39. LiuX. YeK. van VlijmenH.W.T. IJzermanA.P. van WestenG.J.P. DrugEx v3: Scaffold-constrained drug design with graph transformer-based reinforcement learning.J. Cheminform.20231512410.1186/s13321‑023‑00694‑z 36803659
     [Google Scholar]
  40. StåhlN. FalkmanG. KarlssonA. MathiasonG. BoströmJ. Deep reinforcement learning for multiparameter optimization in de novo drug design.J. Chem. Inf. Model.20195973166317610.1021/acs.jcim.9b00325 31273995
     [Google Scholar]
  41. PolykovskiyD. ZhebrakA. VetrovD. IvanenkovY. AladinskiyV. MamoshinaP. BozdaganyanM. AliperA. ZhavoronkovA. KadurinA. Entangled conditional adversarial autoencoder for de novo drug discovery.Mol. Pharm.201815104398440510.1021/acs.molpharmaceut.8b00839
     [Google Scholar]
  42. De CaoN. KipfT. MolGAN: An implicit generative model for small molecular graphs.arXiv2018201811973
     [Google Scholar]
  43. GuimaraesG.L. Sanchez-LengelingB. OuteiralC. FariasP.L.C. Aspuru-GuzikA.n. Objective-reinforced generative adversarial networks (organ) for sequence generation models.arXiv2017201710843
     [Google Scholar]
  44. IgashovI. StärkH. VignacC. SatorrasV.G. FrossardP. WellingM. BronsteinM. CorreiaB. Equivariant 3dconditional diffusion models for molecular linker design.arXiv2022202205274
     [Google Scholar]
  45. IrwinJ.J. ShoichetB.K. ZINC--a free database of commercially available compounds for virtual screening.J. Chem. Inf. Model.200545117718210.1021/ci049714+ 15667143
     [Google Scholar]
  46. LetovskyS. CottinghamR.W. PorterC.J. LiP.W.D. GDB: The human genome database.Nucleic Acids Res.1998261949910.1093/nar/26.1.94 9399808
     [Google Scholar]
  47. GaultonA. BellisL.J. BentoA.P. ChambersJ. DaviesM. HerseyA. LightY. McGlincheyS. MichalovichD. Al-LazikaniB. OveringtonJ.P. ChEMBL: A large-scale bioactivity database for drug discovery.Nucleic Acids Res.201240D1D1100D110710.1093/nar/gkr777 21948594
     [Google Scholar]
  48. RuddigkeitL. van DeursenR. BlumL.C. ReymondJ.L. Enumeration of 166 billion organic small molecules in the chemical universe database GDB-17.J. Chem. Inf. Model.201252112864287510.1021/ci300415d 23088335
     [Google Scholar]
  49. IrwinJ.J. TangK.G. YoungJ. DandarchuluunC. WongB.R. KhurelbaatarM. MorozY.S. MayfieldJ. SayleR.A. ZINC20-a free ultralarge-scale chemical database for ligand discovery.J. Chem. Inf. Model.202060126065607310.1021/acs.jcim.0c00675 33118813
     [Google Scholar]
  50. KimS. ThiessenP.A. BoltonE.E. ChenJ. FuG. GindulyteA. HanL. HeJ. HeS. ShoemakerB.A. WangJ. YuB. ZhangJ. BryantS.H. PubChem substance and compound databases.Nucleic Acids Res.201644D1D1202D121310.1093/nar/gkv951 26400175
     [Google Scholar]
  51. WishartD.S. KnoxC. GuoA.C. ShrivastavaS. HassanaliM. StothardP. ChangZ. WoolseyJ. DrugBank: A comprehensive resource for in silico drug discovery and exploration.Nucleic Acids Res.20063490001D668D67210.1093/nar/gkj067 16381955
     [Google Scholar]
  52. RamakrishnanR. DralP.O. RuppM. von LilienfeldO.A. Quantum chemistry structures and properties of 134 kilo molecules.Sci. Data20141114002210.1038/sdata.2014.22 25977779
     [Google Scholar]
  53. MysingerM.M. CarchiaM. IrwinJ.J. ShoichetB.K. Directory of useful decoys, enhanced (DUD-E): Better ligands and decoys for better benchmarking.J. Med. Chem.201255146582659410.1021/jm300687e 22716043
     [Google Scholar]
  54. WeiningerD. WeiningerA. WeiningerJ.L. SMILES. 2. Algorithm for generation of unique SMILES notation.J. Chem. Inf. Comput. Sci.19892929710110.1021/ci00062a008
     [Google Scholar]
  55. SeglerM.H.S. KogejT. TyrchanC. WallerM.P. Generating focused molecule libraries for drug discovery with recurrent neural networks.ACS Cent. Sci.20184112013110.1021/acscentsci.7b00512 29392184
     [Google Scholar]
  56. KadurinA. NikolenkoS. KhrabrovK. AliperA. ZhavoronkovA. druGAN: An advanced generative adversarial autoencoder model for de novo generation of new molecules with desired molecular properties in silico.Mol. Pharm.20171493098310410.1021/acs.molpharmaceut.7b00346 28703000
     [Google Scholar]
  57. GohG.B. SiegelC. VishnuA. HodasN.O. BakerN. Chemception: A deep neural network with minimal chemistry knowledge matches the performance of expert-developed QSAR/QSPR models.arXiv2017201706689
     [Google Scholar]
  58. LiY. PanQ. WangS. PengH. YangT. CambriaE. Disentangled variational auto-encoder for semi-supervised learning.Inf. Sci.2019482738510.1016/j.ins.2018.12.057
     [Google Scholar]
  59. WangW. WangY. ZhaoH. SciabolaS. A transformerbased generative model for de novo molecular design.arXiv2022202208749
     [Google Scholar]
  60. ZarembaW. SutskeverI. VinyalsO. Recurrent neural network regularization. arXiv201420142329
     [Google Scholar]
  61. HochreiterS. SchmidhuberJ. Long short-term memory.Neural Comput.1997981735178010.1162/neco.1997.9.8.1735 9377276
     [Google Scholar]
  62. ChoK. Learning phrase representations using RNN encoderdecoder for statistical machine translation.arXiv201420141078
     [Google Scholar]
  63. GoodfellowI. Pouget-AbadieJ. MirzaM. XuB. Warde-FarleyD. OzairS. CourvilleA. BengioY. Generative adversarial networks.Commun. ACM2020631113914410.1145/3422622
     [Google Scholar]
  64. WilliamsR.J. Simple statistical gradient-following algorithms for connectionist reinforcement learning.Mach. Learn.199283-422925610.1007/BF00992696
     [Google Scholar]
  65. RanzatoM.A. ChopraS. AuliM. ZarembaW. Sequence level training with recurrent neural networks.arXiv2015201506732
     [Google Scholar]
  66. BrowneC.B. PowleyE. WhitehouseD. LucasS.M. CowlingP.I. RohlfshagenP. TavenerS. PerezD. SamothrakisS. ColtonS. A survey of monte carlo tree search methods.IEEE Trans. Comput. Intell. AI Games20124114310.1109/TCIAIG.2012.2186810
     [Google Scholar]
  67. MorleyS.D. AfsharM. Validation of an empirical RNA-ligand scoring function for fast flexible docking using RiboDock®.J. Comput. Aided Mol. Des.200418318920810.1023/B:JCAM.0000035199.48747.1e 15368919
     [Google Scholar]
  68. Ruiz-CarmonaS. Alvarez-GarciaD. FoloppeN. Garmendia-DovalA.B. JuhosS. SchmidtkeP. BarrilX. HubbardR.E. MorleyS.D. rDock: A fast, versatile and open source program for docking ligands to proteins and nucleic acids.PLOS Comput. Biol.2014104e100357110.1371/journal.pcbi.1003571 24722481
     [Google Scholar]
  69. LipinskiC.A. LombardoF. DominyB.W. FeeneyP.J. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings.Adv. Drug Deliv. Rev.2001461-332610.1016/S0169‑409X(00)00129‑0 11259830
     [Google Scholar]
  70. PolykovskiyD. ZhebrakA. Sanchez-LengelingB. GolovanovS. TatanovO. BelyaevS. KurbanovR. ArtamonovA. AladinskiyV. VeselovM. KadurinA. JohanssonS. ChenH. NikolenkoS. Aspuru-GuzikA. ZhavoronkovA. Molecular sets (MOSES): A benchmarking platform for molecular generation models.Front. Pharmacol.20201156564410.3389/fphar.2020.565644 33390943
     [Google Scholar]
  71. BenhendaM. ChemGAN challenge for drug discovery: Can AI reproduce natural chemical diversity?arXiv2017201708227
     [Google Scholar]
  72. BrownN. FiscatoM. SeglerM.H.S. VaucherA.C. GuacaMol: Benchmarking models for de novo molecular design.J. Chem. Inf. Model.20195931096110810.1021/acs.jcim.8b00839 30887799
     [Google Scholar]
  73. RamsundarB. EastmanP. WaltersP. PandeV. LeswingK. WuZ. Deep Learning for the Life Sciences.O'Reilly Media2019
     [Google Scholar]
  74. ZhuZ. ShiC. ZhangZ. LiuS. XuM. YuanX. ZhangY. ChenJ. CaiH. LuJ. Torchdrug: A powerful and flexible machine learning platform for drug discovery.arXiv2022202208320
     [Google Scholar]
  75. LamH.T. SbodioM.L. GallindoM.M. ZayatsM. Fernández-DíazR. VallsV. PiccoG. RamisC.B. LópezV. Otter- Knowledge: Benchmarks of multimodal knowledge graph representation learning from different sources for drug discovery.arXiv2023202312802
     [Google Scholar]
  76. GodinezW.J. MaE.J. ChaoA.T. PeiL. Skewes-CoxP. CanhamS.M. JenkinsJ.L. YoungJ.M. MartinE.J. GuiguemdeW.A. Design of potent antimalarials with generative chemistry.Nat. Mach. Intell.20224218018610.1038/s42256‑022‑00448‑w
     [Google Scholar]
  77. YoshikawaN. TerayamaK. SumitaM. HommaT. OonoK. TsudaK. Population-based de novo molecule generation, using grammatical evolution.Chem. Lett.201847111431143410.1246/cl.180665
     [Google Scholar]
/content/journals/cad/10.2174/0115734099287389240126072433
Loading
Status and Prospects of Research on Deep Learning-based De Novo Generation of Drug Molecules
Curr. Computeraided Drug Des. 21, 257 (2025); https://doi.org/10.2174/0115734099287389240126072433
/content/journals/cad/10.2174/0115734099287389240126072433
/content/journals/cad/10.2174/0115734099287389240126072433
Loading

Data & Media loading...


  • Article Type:     Review Article
Keyword(s): Deep learning; diffusion model; drug discovery; encoder-decoder; GAN; reinforcement learning; RNN

Most Cited Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error
Please enter a valid_number test