Optimization of Molecules via Deep Reinforcement Learning

Abstract

We present a framework, which we call Molecule Deep Q -Networks (MolDQN), for molecule optimization by combining domain knowledge of chemistry and state-of-the-art reinforcement learning techniques (double Q -learning and randomized value functions). We directly define modifications on molecules, thereby ensuring 100% chemical validity. Further, we operate without pre-training on any dataset to avoid possible bias from the choice of that set. MolDQN achieves comparable or better performance against several other recently published algorithms for benchmark molecular optimization tasks. However, we also argue that many of these tasks are not representative of real optimization problems in drug discovery. Inspired by problems faced during medicinal chemistry lead optimization, we extend our model with multi-objective reinforcement learning, which maximizes drug-likeness while maintaining similarity to the original molecule. We further show the path through chemical space to achieve optimization for a molecule to understand how the model works.

Full text

Its licence does not allow us to show the full text. Read it at PubMed Central:

References 36

1. 1.

   Hughes JP, Rees S, Kalindjian SB, Philpott KL. Principles of early drug discovery. Br. J. Pharmacol. 2011;162:1239-1249.

   Not in our library10.1111/j.1476-5381.2010.01127.x

2. 2.
   Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules · 2018

   Not checked
3. 3.
   Application of Generative Autoencoder in De Novo Molecular Design · 2017

   No notice
4. 4.

   Segler MH, Kogej T, Tyrchan C, Waller MP. Generating focused molecule libraries for drug discovery with recurrent neural networks. ACS Cent. Sci. 2017;4:120-131.

   Not in our library10.1021/acscentsci.7b00512

5. 5.
   Molecular generative model based on conditional variational autoencoder for de novo molecular design · 2018

   No notice
6. 6.

   Putin E, et al. Adversarial threshold neural computer for molecular de novo design. Mol. Pharm. 2018;15:4386-4397.

   Not in our library10.1021/acs.molpharmaceut.7b01137

7. 7.

   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.

   Not in our library10.1021/ci00057a005

8. 8.

   Kusner, M. J., Paige, B. & Hernández-Lobato, J. M. Grammar variational autoencoder. arXiv preprint arXiv:1703.01925 (2017).

   Not in our library

9. 9.

   Dai, H., Tian, Y., Dai, B., Skiena, S. & Song, L. Syntax-directed variational autoencoder for structured data. arXiv preprint arXiv:1802.08786 (2018).

   Not in our library

10. 10.

    Li, Y., Vinyals, O., Dyer, C., Pascanu, R. & Battaglia, P. Learning deep generative models of graphs. arXiv preprint arXiv:1803.03324 (2018).

    Not in our library

11. 11.
    Multi-objective de novo drug design with conditional graph generative model · 2018

    No notice
12. 12.

    Cao, N. De & Kipf, T. Molgan: An implicit generative model for small molecular graphs. arXiv preprint arXiv:1805.11973 (2018).

    Not in our library

13. 13.

    Jin, W., Barzilay, R. & Jaakkola, T. Junction tree variational autoencoder for molecular graph generation. arXiv preprint arXiv:1802.04364 (2018).

    Not in our library

14. 14.
    Molecular de-novo design through deep reinforcement learning · 2017

    No notice
15. 15.

    Guimaraes, G. L., Sanchez-Lengeling, B., Outeiral, C., Farias, P. L. C. & Aspuru-Guzik, A. Objective-reinforced generative adversarial networks (organ) for sequence generation models. arXiv preprint arXiv:1705.10843 (2017).

    Not in our library

16. 16.

    Putin, E. et al. Reinforced adversarial neural computer for de novo molecular design. J. Chem. Inf. Model. (2018).

    Not in our library10.1021/acs.jcim.7b00690

17. 17.

    Popova M, Isayev O, Tropsha A. Deep reinforcement learning for de novo drug design. Sci. Adv. 2018;4:eaap7885.

    Not in our library10.1126/sciadv.aap7885

18. 18.

    You, J., Liu, B., Ying, R., Pande, V. & Leskovec, J. Graph convolutional policy network for goal-directed molecular graph generation. arXiv preprint arXiv:1806.02473 (2018).

    Not in our library

19. 19.

    Bellman Richard. A Markovian Decision Process. Indiana University Mathematics Journal. 1957;6(4):679-684.

    Not in our library10.1512/iumj.1957.6.56038

20. 20.

    Mnih V, et al. Human-level control through deep reinforcement learning. Nature. 2015;518:529.

    Not in our library10.1038/nature14236

21. 21.

    Gu, S., Lillicrap, T., Ghahramani, Z., Turner, R. E. & Levine, S. Q-prop: Sample-efficient policy gradient with an off-policy critic. arXiv preprint arXiv:1611.02247 (2016).

    Not in our library

22. 22.

    Rdkit. Rdkit: Open-source cheminformatics software, http://www.rdkit.org/, https://github.com/rdkit/rdkit (2016).

    Not in our library

23. 23.

    Boyd, S. & Vandenberghe, L. Convex optimization (Cambridge university press, 2004).

    Not in our library

24. 24.

    Osband, I., Blundell, C., Pritzel, A. & Van Roy, B. Deep exploration via bootstrapped DQN in Adv. Neural Inf. Process. Syst. 4026–4034 (2016).

    Not in our library

25. 25.

    Van Hasselt, H., Guez, A. & Silver, D. Deep reinforcement learning with double Q-Learning. In AAAI2, 5 (2016).

    Not in our library

26. 26.

    Rogers D, Hahn M. Extended-connectivity fingerprints. J. Chem. Inf. Model. 2010;50:742-754.

    Not in our library10.1021/ci100050t

27. 27.

    Kingma, D. P. & Ba, J. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014).

    Not in our library

28. 28.

    Bickerton GR, Paolini GV, Besnard J, Muresan S, Hopkins AL. Quantifying the chemical beauty of drugs. Nat. Chem. 2012;4:90.

    Not in our library10.1038/nchem.1243

29. 29.

    Irwin JJ, Sterling T, Mysinger MM, Bolstad ES, Coleman RG. Zinc: a free tool to discover chemistry for biology. J. Chem. Inf. Model. 2012;52:1757-1768.

    Not in our library10.1021/ci3001277

30. 30.

    Welch BL. The generalization ofstudent’s’ problem when several different population variances are involved. Biometrika. 1947;34:28-35.

    Not in our library10.1093/biomet/34.1-2.28

31. 31.

    Garg, T., Singh, O., Arora, S. & Murthy, R. Scaffold: a novel carrier for cell and drug delivery, Crit. Rev. Ther. Drug. Carrier. Syst. 29 (2012).

    Not in our library10.1615/critrevtherdrugcarriersyst.v29.i1.10

32. 32.
    The ChEMBL database in 2017 · 2016

    No notice
33. 33.

    Haarnoja, T., Tang, H., Abbeel, P. & Levine, S. Reinforcement learning with deep energy-based policies. arXiv preprint arXiv:1702.08165 (2017).

    Not in our library

34. 34.

    Gilmer, J., Schoenholz, S. S., Riley, P. F., Vinyals, O. & Dahl, G. E. Neural message passing for quantum chemistry. arXiv preprint arXiv:1704.01212 (2017).

    Not in our library

35. 35.

    Merk D, Friedrich L, Grisoni F, Schneider G. De novo design of bioactive small molecules by artificial intelligence. Mol. Inform. 2018;37:1700153.

    Not in our library10.1002/minf.201700153

36. 36.

    Benhenda, M. Chemgan challenge for drug discovery: can ai reproduce natural chemical diversity? arXiv preprint arXiv:1708.08227 (2017).

    Not in our library

Cited by papers we hold 2

• Generative Models as an Emerging Paradigm in the Chemical Sciences · 2023
  No notice
• Constrained Bayesian optimization for automatic chemical design using variational autoencoders · 2020
  No notice

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.