Training recurrent neural networks as generative neural networks for molecular structures: how does it impact drug discovery?

References

• Schneider G. Automating drug discovery. Nat Rev Drug Discov. 2018;17:97–113.
   PubMed Web of Science ®Google Scholar
• Waring MJ, Arrowsmith J, Leach AR, et al. An analysis of the attrition of drug candidates from four major pharmaceutical companies. Nat Rev Drug Discov. 2015;14:475–486.
   PubMed Web of Science ®Google Scholar
• Reymond J-L, Ruddigkeit L, Blum L, et al. The enumeration of chemical space. Wiley Interdisci: Rev Comput Mol Sci. 2012;2:717–733.
   Web of Science ®Google Scholar
• Schneider P, Walters WP, Plowright AT, et al. Rethinking drug design in the artificial intelligence era. Nat Rev Drug Discov. 2020;19:353–364.
   PubMed Web of Science ®Google Scholar
• Schneider G, Fechner U. Computer-based de novo design of drug-like molecules. Nat Rev Drug Discov. 2005;4:649–663.
   PubMed Web of Science ®Google Scholar
• Churchwell CJ, Rintoul MD, Martin S, et al. The signature molecular descriptor: 3. inverse-quantitative structure–activity relationship of icam-1 inhibitory peptides. J Mol Graphics Modell. 2004;22:263–273.
   Google Scholar
• Wong WW, Burkowski FJ. A constructive approach for discovering new drug leads: using a kernel methodology for the inverse-QSAR problem. J Cheminform. 2009;1:1–27.
   PubMed Web of Science ®Google Scholar
• Faulon J-L, Churchwell CJ, Visco DP. The signature molecular descriptor. 2. enumerating molecules from their extended valence sequences. J Chem Inf Comput Sci. 2003;43:721–734.
   PubMedGoogle Scholar
• Miyao T, Kaneko H, Funatsu K. Inverse qspr/qsar analysis for chemical structure generation (from y to x. J Chem Inf Model. 2016;56:286–299.
   PubMed Web of Science ®Google Scholar
• Jørgensen PB, Schmidt MN, Winther O. Deep generative models for molecular science. Mol Inf. 2018;37:1700133.
   Web of Science ®Google Scholar
• Brown N, Lewis RA. Exploiting QSAR methods in lead optimization. Curr Opin Drug Discov. Develop. 2006;9:419–424.
   PubMed Web of Science ®Google Scholar
• Langevin M, Minoux H, Levesque M, et al. Scaffold-constrained molecular generation. J Chem Inf Model. 2020;60:5637–5646.
   PubMed Web of Science ®Google Scholar
• Bagal V, Aggarwal R, Vinod P, et al. MolGPT: molecular generation using a transformer-decoder model. J Chem Inf Model. 2021;62:2064–2076.
   PubMed Web of Science ®Google Scholar
• Kim H, Na J, Lee WB. Generative chemical transformer: neural machine learning of molecular geometric structures from chemical language via attention. J Chem Inf Model. 2021;61:5804–5814.
   PubMed Web of Science ®Google Scholar
• Kadurin A, Nikolenko S, Khrabrov K, et al. DruGAN: an advanced generative adversarial autoencoder model for de novo generation of new molecules with desired molecular properties in silico. Mol Pharm. 2017;14:3098–3104.
   PubMed Web of Science ®Google Scholar
• Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative adversarial networks. Commun ACM. 2020;63:139–144.
   Web of Science ®Google Scholar
• Tong X, Liu X, Tan X, et al. Generative models for de novo drug design. J Med Chem. 2021;64:14011–14027.
   PubMed Web of Science ®Google Scholar
• Xiong J, Xiong Z, Chen K, et al. Graph neural networks for automated de novo drug design. Drug Discov Today. 2021;26:1382–1393.
   PubMed Web of Science ®Google Scholar
• Xu Y, Lin K, Wang S, et al. Deep learning for molecular generation. Future Med Chem. 2019;11:567–597 * Deep learning algorithms and optimization strategies. .
   PubMed Web of Science ®Google Scholar
• Wang M, Wang Z, Sun H, et al. Deep learning approaches for de novo drug design: an overview. Curr Opin Struct Biol. 2022;72:135–144.
   PubMed Web of Science ®Google Scholar
• Bjerrum EJ. Smiles enumeration as data augmentation for neural network modeling of molecules. arXiv preprint arXiv:1703.07076. 2017;
   Google Scholar
• Jastrze˛bski S, Les´niak D, Czarnecki WM. Learning to smile (s), arXiv preprint arXiv:1602.06289. 2016.
   Google Scholar
• Weininger D. SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules. J Chem Inf Comput Sci. 1988;28:31–36.
   Google Scholar
• Bjerrum EJ, Threlfall R. Molecular generation with recurrent neural networks (RNNs), arXiv preprint arXiv:1705.04612. 2017.
   Google Scholar
• Ertl P, Lewis R, Martin E, et al. In silico generation of novel, drug-like chemical matter using the LSTM neural network. arXiv preprint arXiv:1712.07449. 2017.
   Google Scholar
• Gupta A, Müller AT, Huisman BJ, et al. Generative recurrent networks for de novo drug design. Mol Inform. 2018;37:1700111.
   PubMed Web of Science ®Google Scholar
• Segler MH, Kogej T, Tyrchan C, et al. Generating focused molecule libraries for drug discovery with recurrent neural networks. ACS central science. 2017;4:120–131 .
   PubMed Web of Science ®Google Scholar
• Jin W, Barzilay R, Jaakkola T. Junction tree variational autoencoder for molecular graph generation, in: International conference on machine learning, PMLR, 2018, pp. 2323–2332.
   Google Scholar
• Bjerrum EJ, Sattarov B. Improving chemical autoencoder latent space and molecular de novo generation diversity with heteroencoders. Biomolecules. 2018;8:131.
   PubMed Web of Science ®Google Scholar
• Kusner MJ, Paige B, Hernández-Lobato JM. Grammar variational autoencoder, in: International conference on machine learning, PMLR, 6-11 August 2017, Sydney, Australia, 2017, pp. 1945–1954.
   Google Scholar
• Krenn M, Häse F, Nigam A, et al. Self-referencing embedded strings (selfies): a 100% robust molecular string representation. Mach Learn: Sci Technol. 2020;1:045024.
   Google Scholar
• Bronstein MM, Bruna J, LeCun Y, et al. Geometric deep learning: going beyond Euclidean data. IEEE Signal Process Mag. 2017;34:18–42.
   Web of Science ®Google Scholar
• Hamilton WL, Ying R, Leskovec J. Representation learning on graphs: methods and applications. arXiv preprint arXiv:1709.05584. 2017;
   Google Scholar
• Samanta B, De A, Ganguly N, et al., Designing random graph models using variational autoencoders with applications to chemical design. arXiv preprint arXiv:1802.05283. 2018;
   Google Scholar
• Li Y, Vinyals O, Dyer C, et al., Learning deep generative models of graphs. arXiv preprint arXiv:1803.03324. 2018;
   Google Scholar
• Li Y, Zhang L, Liu Z. Multi-objective de novo drug design with conditional graph generative model. J Cheminform. 2018;10:1–24.
   PubMed Web of Science ®Google Scholar
• Simonovsky M, Komodakis N. Graphvae: towards generation of small graphs using variational autoencoders, in: 27th International Conference on Artificial Neural Networks, October 4-7, 2018, Rhodes, Greece, Springer, 2018, pp. 412–422.
   Google Scholar
• Samanta B, De A, Jana G, et al. Nevae: a deep generative model for molecular graphs. J Mach Learn Res. 2020 Apr;21(114): 1–33.
   PubMedGoogle Scholar
• Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative adversarial networks, Communications of the ACM 63 (2020). 28th Annual Conference on Neural Information Processing Systems 2014 [(NIPS)]; December 8-13, 2014. Montreal, Canada. 2014.
   Google Scholar
• De Cao N, Kipf T. Molgan: an implicit generative model for small molecular graphs. arXiv preprint arXiv:1805.11973. 2018;
   Google Scholar
• Kwon Y, Lee D, Choi Y-S, et al. Compressed graph representation for scalable molecular graph generation. J Cheminform. 2020;12:1–8.
   PubMed Web of Science ®Google Scholar
• Meyers J, Fabian B, Brown N. De novo molecular design and generative models. Drug Discov Today. 2021;26:2707–2715.
   PubMed Web of Science ®Google Scholar
• Brown N, Fiscato M, Segler MH, et al. Guacamol: benchmarking models for de novo molecular design. J Chem Inf Model. 2019;59:1096–1108.
   PubMed Web of Science ®Google Scholar
• Yang X, Zhang J, Yoshizoe K, et al. Chemts: an efficient python library for de novo molecular generation. Sci Technol Adv Mater. 2017;18:972–976.
   PubMed Web of Science ®Google Scholar
• Dai H, Tian Y, Dai B, et al. Syntax-directed variational autoencoder for structured data, arXiv preprint arXiv:1802.08786. 2018.
   Google Scholar
• Blaschke T, Olivecrona M, Engkvist O, et al. Application of generative autoencoder in de novo molecular design. Mol Inform. 2018;37:1700123.
   PubMed Web of Science ®Google Scholar
• Polykovskiy D, Zhebrak A, Vetrov D, et al. Entangled conditional adversarial autoencoder for de novo drug discovery. Mol Pharm. 2018;15:4398–4405.
   PubMed Web of Science ®Google Scholar
• Guimaraes GL, Sanchez-Lengeling B, Outeiral C, et al., Objective-reinforced gen erative adversarial networks (ORGAN) for sequence generation models. arXiv preprint arXiv:1705.10843. 2017;
   Google Scholar
• Sanchez-Lengeling B, Outeiral C, Guimaraes GL, et al. Optimizing distributions over molecular space. An objective-reinforced generative adversarial network for inverse-design chemistry (organic). ChemRxiv. 2017. DOI: 10.26434/chemrxiv.5309668.v3.
   Google Scholar
• Putin E, Asadulaev A, Vanhaelen Q, et al. Adversarial threshold neural computer for molecular de novo design. Mol Pharm. 2018;15:4386–4397.
   PubMedGoogle Scholar
• Putin E, Asadulaev A, Ivanenkov Y, et al. Reinforced adversarial neural computer for de novo molecular design. J Chem Inf Model. 2018;58:1194–1204.
   PubMed Web of Science ®Google Scholar
• Liu Q, Allamanis M, Brockschmidt M, et al. Constrained graph variational autoencoders for molecule design. Adv Neural Inf Process Syst. 2018;31.
   Google Scholar
• You J, Liu B, Ying Z, et al. Graph convolutional policy network for goal-directed molecular graph generation. NIPS'18: Proceedings of the 32nd International Conference on Neural Information Processing Systems. December 3-8, 2018, Montreal Canada. 2018;31.
   Google Scholar
• Liu X. Deep recurrent neural network for protein function prediction from sequence. arXiv preprint arXiv:1701.08318. 2017;
   Google Scholar
• Lusci A, Pollastri G, Baldi P. Deep architectures and deep learning in chemoinformatics: the prediction of aqueous solubility for drug-like molecules. J Chem Inf Model. 2013;53:1563–1575.
   PubMed Web of Science ®Google Scholar
• Gómez-Bombarelli R, Wei JN, Duvenaud D, et al. Automatic chemical design using a data- driven continuous representation of molecules. ACS central science. 2018;4:268–276.
   PubMed Web of Science ®Google Scholar
• Hochreiter S, Schmidhuber J. Long short-term memory. Neural computation. 1997;9:1735–1780.
   PubMed Web of Science ®Google Scholar
• Cho K, Van Merriënboer B, Gulcehre C, et al., Learning phrase representations using rnn encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078. 2014;
   Google Scholar
• Segler MH, Kogej T, Tyrchan C, et al. Generating focused molecule libraries for drug discovery with recurrent neural networks. ACS Centr Sci. 2018;4:120–131.
   PubMed Web of Science ®Google Scholar
• Merk D, Friedrich L, Grisoni F, et al. De novo design of bioactive small molecules by artificial intelligence. Mol Inform. 2018;37:1700153.
   PubMed Web of Science ®Google Scholar
• Grisoni F, Moret M, Lingwood R, et al. Bidirectional molecule generation with recurrent neural networks. J Chem Inf Model. 2020;60:1175–1183.
   PubMed Web of Science ®Google Scholar
• Arús-Pous J, Johansson SV, Prykhodko O, et al. Randomized smiles strings improve the quality of molecular generative models. J Cheminform. 2019;11:1–13.
   PubMed Web of Science ®Google Scholar
• Popova M, Isayev O, Tropsha A. Deep reinforcement learning for de novo drug design. Sci Adv. 2018;4: eaap7885.
   PubMed Web of Science ®Google Scholar
• Yu L, Zhang W, Wang J, et al., Seqgan: sequence generative adversarial nets with policy gradient, in: Proceedings of the AAAI conference on artificial intelligence, San Francisco, California USA, volume 31, 2017.
   Google Scholar
• Jaques N, Gu S, Bahdanau D, et al., Sequence tutor: conservative fine-tuning of sequence generation models with kl-control, in: Proceedings of the 34th International Conference on Machine Learning (pp. 1645-1654). PMLR. Sydney, Australia, 2017, pp. 1645–1654.
   Google Scholar
• Leo A, Hansch C, Elkins D. Partition coefficients and their uses. Chem Rev. 1971;71:525–616.
   Web of Science ®Google Scholar
• Bickerton GR, Paolini GV, Besnard J, et al. Quantifying the chemical beauty of drugs. Nat Chem. 2012;4:90–98.
   PubMed Web of Science ®Google Scholar
• Olivecrona M, Blaschke T, Engkvist O, et al. Molecular de-novo design through deep reinforcement learning. J Cheminform. 2017;9:1–14.
   PubMed Web of Science ®Google Scholar
• Cires¸an DC, Meier U, Schmidhuber J. Transfer learning for Latin and Chinese characters with deep neural networks, in: The 2012 international joint conference on neural networks (IJCNN) Brisbane, Australia, IEEE, 2012, pp. 1–6.
   Google Scholar
• Pan SJ, Yang Q. A survey on transfer learning. IEEE Trans Knowledge Data Eng. 2009;22:1345–1359.
   Web of Science ®Google Scholar
• Torrey L, Shavlik J. Transfer learning. In: Handbook of research on machine learning applications and trends: algorithms, methods, and techniques, IGI global. Hershey, PA: Information Science Reference - Imprint of: IGI Publishing. 2010. p. 242–264.
   Google Scholar
• Fonseca CM, Fleming PJ. Genetic algorithms for multiobjective optimization: formulation discussion and generalization. In: ICGA. 1993. Vol. 93. Citeseer. p. 416–423.
   Google Scholar
• Jhoti H, Williams G, Rees DC, et al. The’rule of three’for fragment-based drug discovery: where are we now? Nat Rev Drug Discov. 2013;12: 644–644.
   PubMed Web of Science ®Google Scholar
• Li X, Xu Y, Yao H, et al. Chemical space exploration based on recurrent neural networks: applications in discovering kinase inhibitors. J Cheminform. 2020;12:1–13.
   PubMed Web of Science ®Google Scholar
• Renz P, Van Rompaey D, Wegner JK, et al. On failure modes in molecule generation and optimization. Drug Discov Today. 2019;32:55–63.
   Google Scholar
• Preuer K, Renz P, Unterthiner T, et al. Fréchet chemnet distance: a metric for generative models for molecules in drug discovery. J Chem Inf Model. 2018;58:1736–1741.
   PubMed Web of Science ®Google Scholar
• Polykovskiy D, Zhebrak A, Sanchez-Lengeling B, et al. Molecular sets (MOSES): a benchmarking platform for molecular generation models. Front Pharmacol. 2020;11:565644.
   PubMed Web of Science ®Google Scholar
• Amabilino S, Pogány P, Pickett SD, et al. Guidelines for recurrent neural network transfer learning- based molecular generation of focused libraries. J Chem Inf Model. 2020;60:5699–5713.
   PubMed Web of Science ®Google Scholar