GEOM, energy-annotated molecular conformations for property prediction and molecular generation

Abstract

Machine learning (ML) outperforms traditional approaches in many molecular design tasks. ML models usually predict molecular properties from a 2D chemical graph or a single 3D structure, but neither of these representations accounts for the ensemble of 3D conformers that are accessible to a molecule. Property prediction could be improved by using conformer ensembles as input, but there is no large-scale dataset that contains graphs annotated with accurate conformers and experimental data. Here we use advanced sampling and semi-empirical density functional theory (DFT) to generate 37 million molecular conformations for over 450,000 molecules. The Geometric Ensemble Of Molecules (GEOM) dataset contains conformers for 133,000 species from QM9, and 317,000 species with experimental data related to biophysics, physiology, and physical chemistry. Ensembles of 1,511 species with BACE-1 inhibition data are also labeled with high-quality DFT free energies in an implicit water solvent, and 534 ensembles are further optimized with DFT. GEOM will assist in the development of models that predict properties from conformer ensembles, and generative models that sample 3D conformations.

Full text

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

References 107

1. 1.

   Norinder U, Lidén P, Boström H. Discrimination between modes of toxic action of phenols using rule based methods. Molecular diversity. 2006;10:207-212.

   Not in our library10.1007/s11030-006-9019-3

2. 2.
   Molecular dynamics simulations and drug discovery · 2011

   No notice
3. 3.

   Stokes JM, et al. A deep learning approach to antibiotic discovery. Cell. 2020;180:688-702.

   Not in our library10.1016/j.cell.2020.01.021

4. 4.

   Gómez-Bombarelli R, et al. Design of efficient molecular organic light-emitting diodes by a high-throughput virtual screening and experimental approach. Nature Materials. 2016;15:1120-1127.

   Not in our library10.1038/nmat4717

5. 5.

   Zhavoronkov A, et al. Deep learning enables rapid identification of potent DDR1 kinase inhibitors. Nature biotechnology. 2019;37:1038-1040.

   Not in our library10.1038/s41587-019-0224-x

6. 6.

   Schwalbe-Koda, D. & Gómez-Bombarelli, R. Generative models for automatic chemical design. In Machine Learning Meets Quantum Physics, 445–467 10.1007/978-3-030-40245-7_21 (Springer, 2020).

   Not in our library

7. 7.
   Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules · 2018

   Not checked
8. 8.

   Jin, W., Barzilay, R. & Jaakkola, T. Junction Tree Variational Autoencoder for Molecular Graph Generation. In International Conference on Machine Learning, https://proceedings.mlr.press/v80/jin18a.html (2018).

   Not in our library

9. 9.

   Dai, H., Tian, Y., Dai, B., Skiena, S. & Song, L. Syntax-directed variational autoencoder for structured data. In International Conference on Learning Representations, https://openreview.net/forum?id=SyqShMZRb (2018).

   Not in our library

10. 10.

    Noé F, Olsson S, Köhler J, Wu H. Boltzmann generators: Sampling equilibrium states of many-body systems with deep. Science. 2019;365:eaaw1147.

    Not in our library10.1126/science.aaw1147

11. 11.
    Molecular de-novo design through deep reinforcement learning · 2017

    No notice
12. 12.

    Gottipati, S. K. et al. Learning to navigate the synthetically accessible chemical space using reinforcement learning. In International Conference on Machine Learning, 3668–3679, https://proceedings.mlr.press/v119/gottipati20a.html (PMLR, 2020)

    Not in our library

13. 13.

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

    Not in our library10.1126/sciadv.aap7885

14. 14.

    AlQuraishi M. End-to-end differentiable learning of protein structure. Cell Systems. 2019;8:292-301.e3.

    Not in our library10.1016/j.cels.2019.03.006

15. 15.

    Ingraham, J., Riesselman, A., Sander, C. & Marks, D. Learning protein structure with a differentiable simulator. In International Conference on Learning Representations, https://openreview.net/forum?id=Byg3y3C9Km (2019).

    Not in our library

16. 16.

    Segler MHS, Preuss M, Waller MP. Planning chemical syntheses with deep neural networks and symbolic AI. Nature. 2018;555:604-610.

    Not in our library10.1038/nature25978

17. 17.

    Coley CW, Barzilay R, Jaakkola TS, Green WH, Jensen KF. Prediction of organic reaction outcomes using machine learning. ACS Central Science. 2017;3:434-443.

    Not in our library10.1021/acscentsci.7b00064

18. 18.

    Duvenaud, D. K. et al. Convolutional networks on graphs for learning molecular fingerprints. In Advances in Neural Information Processing Systems, 2215–2223, https://proceedings.neurips.cc/paper/2015/file/f9be311e65d81a9ad8150a60844bb94c-Paper.pdf (2015).

    Not in our library

19. 19.

    Kearnes S, McCloskey K, Berndl M, Pande V, Riley P. Molecular graph convolutions: moving beyond fingerprints. Journal of Computer-Aided Molecular Design. 2016;30:595-608.

    Not in our library10.1007/s10822-016-9938-8

20. 20.

    Yang K, et al. Analyzing learned molecular representations for property prediction. Journal of Chemical Information and Modeling. 2019;59:3370-3388.

    Not in our library10.1021/acs.jcim.9b00237

21. 21.

    Anderson, B., Hy, T. S. & Kondor, R. Cormorant: Covariant molecular neural networks. In Advances in Neural Information Processing Systems, 14537–14546, https://proceedings.neurips.cc/paper/2019/file/03573b32b2746e6e8ca98b9123f2249b-Paper.pdf (2019).

    Not in our library

22. 22.

    Klicpera, J., Groß, J. & Günnemann, S. Directional message passing for molecular graphs. In International Conference on Learning Representations, https://openreview.net/forum?id=B1eWbxStPH (2019).

    Not in our library

23. 23.

    Ramsundar, B. et al. Deep Learning for the Life Sciences (O’Reilly Media, 2019).

    Not in our library

24. 24.
    ChEMBL: towards direct deposition of bioassay data · 2018

    No notice
25. 25.

    Sterling T, Irwin JJ. ZINC 15–Ligand discovery for everyone. Journal of chemical information and modeling. 2015;55:2324-37.

    Not in our library10.1021/acs.jcim.5b00559

26. 26.

    Brown N, Fiscato M, Segler MHS, Vaucher AC. GuacaMol: Benchmarking models for de novo molecular design. J. Chem. Inf. Model. 2019;59:1096-1108.

    Not in our library10.1021/acs.jcim.8b00839

27. 27.
    Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models · 2020

    No notice
28. 28.

    Delaney JS. ESOL: Estimating aqueous solubility directly from molecular structure. Journal of Chemical Information and Computer Sciences. 2004;44:1000-1005.

    Not in our library10.1021/ci034243x

29. 29.

    Mobley DL, Guthrie JP. FreeSolv: A database of experimental and calculated hydration free energies, with input files. Journal of Computer-Aided Molecular Design. 2014;28:711-720.

    Not in our library10.1007/s10822-014-9747-x

30. 30.

    Wang R, Fang X, Lu Y, Wang S. The PDBbind database: Collection of binding affinities for protein-ligand complexes with known three-dimensional structures. Journal of Medicinal Chemistry. 2004;47:2977-2980.

    Not in our library10.1021/jm030580l

31. 31.

    Wu Z, et al. MoleculeNet: a benchmark for molecular machine learning. Chemical science. 2018;9:513-530.

    Not in our library10.1039/c7sc02664a

32. 32.

    Weininger D. SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules. Journal of Chemical Information and Modeling. 1988;28:31-36.

    Not in our library10.1021/ci00057a005

33. 33.
    InChI, the IUPAC International Chemical Identifier · 2015

    No notice
34. 34.

    Kuhn B, et al. A real-world perspective on molecular design: Miniperspective. Journal of medicinal chemistry. 2016;59:4087-4102.

    Not in our library10.1021/acs.jmedchem.5b01875

35. 35.

    Hawkins PC. Conformation generation: The state of the art. Journal of chemical information and modeling. 2017;57:1747-1756.

    Not in our library10.1021/acs.jcim.7b00221

36. 36.

    Ramakrishnan R, Dral PO, Rupp M, von Lilienfeld OA. Quantum chemistry structures and properties of 134 kilo molecules. Scientific data. 2014;1:140022.

    Not in our library10.1038/sdata.2014.22

37. 37.

    Pracht P, Bohle F, Grimme S. Automated exploration of the low-energy chemical space with fast quantum chemical methods. Physical Chemistry Chemical Physics. 2020;22:7169-7192.

    Not in our library10.1039/c9cp06869d

38. 38.

    Bannwarth C, Ehlert S, Grimme S. GFN2-xTB—An accurate and broadly parametrized self-consistent tight-binding quantum chemical method with multipole electrostatics and density-dependent dispersion contributions. Journal of chemical theory and computation. 2019;15:1652-1671.

    Not in our library10.1021/acs.jctc.8b01176

39. 39.

    Subramanian G, Ramsundar B, Pande V, Denny RA. Computational modeling of β-secretase 1 (BACE-1) inhibitors using ligand based approaches. Journal of chemical information and modeling. 2016;56:1936-1949.

    Not in our library10.1021/acs.jcim.6b00290

40. 40.
    Crystallography Open Database – an open-access collection of crystal structures · 2009

    No notice
41. 41.
    The Cambridge Structural Database · 2016

    No notice
42. 42.
    ANI-1, A data set of 20 million calculated off-equilibrium conformations for organic molecules · 2017

    No notice
43. 43.

    Smith JS, Isayev O, Roitberg AE. ANI-1: an extensible neural network potential with DFT accuracy at force field computational cost. Chemical Science. 2017;8:3192-3203.

    Not in our library10.1039/c6sc05720a

44. 44.

    Smith JS, Nebgen B, Lubbers N, Isayev O, Roitberg AE. Less is more: Sampling chemical space with active learning. Journal of Chemical Physics. 2018;148:241733.

    Not in our library10.1063/1.5023802

45. 45.

    Chmiela S, et al. Machine learning of accurate energy-conserving molecular force fields. Science Advances. 2017;3:e1603015.

    Not in our library10.1126/sciadv.1603015

46. 46.

    Simm, G. & Hernandez-Lobato, J. M. A generative model for molecular distance geometry. In International Conference on Machine Learning, 8949–8958, https://proceedings.mlr.press/v119/simm20a.html (PMLR, 2020).

    Not in our library

47. 47.

    Kanal IY, Keith JA, Hutchison GR. A sobering assessment of small-molecule force field methods for low energy conformer predictions. International Journal of Quantum Chemistry. 2018;118:e25512.

    Not in our library10.1002/qua.25512

48. 48.

    Bolton EE, Kim S, Bryant SH. PubChem3D: conformer generation. Journal of cheminformatics. 2011;3:4.

    Not in our library10.1186/1758-2946-3-4

49. 49.

    Simm, G., Pinsler, R. & Hernández-Lobato, J. M. Reinforcement learning for molecular design guided by quantum mechanics. In International Conference on Machine Learning, 8959–8969 https://proceedings.mlr.press/v119/simm20b.html (PMLR, 2020).

    Not in our library

50. 50.

    Stieffenhofer M, Wand M, Bereau T. Adversarial reverse mapping of equilibrated condensed-phase molecular structures. Machine Learning: Science and Technology. 2020;1:045014.

    Not in our library10.1088/2632-2153/abb6d4

51. 51.

    Imrie F, Bradley AR, van der Schaar M, Deane CM. Deep generative models for 3D linker design. Journal of chemical information and modeling. 2020;60:1983-1995.

    Not in our library10.1021/acs.jcim.9b01120

52. 52.

    Mansimov E, Mahmood O, Kang S, Cho K. Molecular geometry prediction using a deep generative graph neural network. Scientific Reports. 2019;9:1-13.

    Not in our library10.1038/s41598-019-56773-5

53. 53.

    Chan L, Hutchison GR, Morris GM. Bayesian optimization for conformer generation. Journal of Cheminformatics. 2019;11:32.

    Not in our library10.1186/s13321-019-0354-7

54. 54.

    Gebauer, N., Gastegger, M. & Schütt, K. Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules. In Advances in neural information processing systems, 32, https://proceedings.neurips.cc/paper/2019/file/a4d8e2a7e0d0c102339f97716d2fdfb6-Paper.pdf (2019).

    Not in our library

55. 55.

    Wang W, Gómez-Bombarelli R. Coarse-graining auto-encoders for molecular dynamics. npj Computational Materials. 2019;5:125.

    Not in our library10.1038/s41524-019-0261-5

56. 56.

    Engel, D. qHTS of yeast-based assay for SARS-CoV PLP. https://pubchem.ncbi.nlm.nih.gov/bioassay/485353.

    Not in our library

57. 57.

    Engel, D. qHTS of yeast-based assay for SARS-CoV PLP: Hit validation. https://pubchem.ncbi.nlm.nih.gov/bioassay/652038.

    Not in our library

58. 58.

    Vainio MJ, Johnson MS. Generating conformer ensembles using a multiobjective genetic algorithm. Journal of chemical information and modeling. 2007;47:2462-2474.

    Not in our library10.1021/ci6005646

59. 59.

    Puranen JS, Vainio MJ, Johnson MS. Accurate conformation-dependent molecular electrostatic potentials for high-throughput in silico drug discovery. Journal of computational chemistry. 2010;31:1722-1732.

    Not in our library10.1002/jcc.21460

60. 60.
    Confab - Systematic generation of diverse low-energy conformers · 2011

    No notice
61. 61.

    Miteva MA, Guyon F, Pierre T. Frog2: Efficient 3D conformation ensemble generator for small compounds. Nucleic acids research. 2010;38:W622-W627.

    Not in our library10.1093/nar/gkq325

62. 62.

    Vilar S, Cozza G, Stefano M. Medicinal chemistry and the molecular operating environment (MOE): application of QSAR and molecular docking to drug discovery. Current topics in medicinal chemistry. 2008;8:1555-1572.

    Not in our library10.2174/156802608786786624

63. 63.

    Hawkins PC, Skillman AG, Warren GL, Ellingson BA, Stahl MT. Conformer generation with OMEGA: algorithm and validation using high quality structures from the Protein Databank and Cambridge Structural Database. Journal of chemical information and modeling. 2010;50:572-584.

    Not in our library10.1021/ci100031x

64. 64.

    RDKit: Open-source cheminformatics. http://www.rdkit.org.

    Not in our library

65. 65.

    Chan L, Hutchison GR, Morris GM. Bayesian optimization for conformer generation. Journal of cheminformatics. 2019;11:1-11.

    Not in our library10.1186/s13321-019-0354-7

66. 66.

    Schwab CH. Conformations and 3D pharmacophore searching. Drug Discovery Today: Technologies. 2010;7:e245-e253.

    Not in our library10.1016/j.ddtec.2010.10.003

67. 67.

    Spellmeyer DC, Wong AK, Bower MJ, Blaney JM. Conformational analysis using distance geometry methods. Journal of Molecular Graphics and Modelling. 1997;15:18-36.

    Not in our library10.1016/s1093-3263(97)00014-4

68. 68.

    Grimme S. Exploration of chemical compound, conformer, and reaction space with meta-dynamics simulations based on tight-binding quantum chemical calculations. Journal of chemical theory and computation. 2019;15:2847-2862.

    Not in our library10.1021/acs.jctc.9b00143

69. 69.

    Grimme S, et al. Fully automated quantum-chemistry-based computation of spin–spin-coupled nuclear magnetic resonance spectra. Angewandte Chemie International Edition. 2017;56:14763-14769.

    Not in our library10.1002/anie.201708266

70. 70.

    Domingos SR, Pérez C, Medcraft C, Pinacho P, Schnell M. Flexibility unleashed in acyclic monoterpenes: Conformational space of citronellal revealed by broadband rotational spectroscopy. Physical Chemistry Chemical Physics. 2016;18:16682-16689.

    Not in our library10.1039/c6cp02876d

71. 71.

    Grimme S, et al. Efficient quantum chemical calculation of structure ensembles and free energies for nonrigid molecules. The Journal of Physical Chemistry A. 2021;125:4039-4054.

    Not in our library10.1021/acs.jpca.1c00971

72. 72.

    Grimme S, Hansen A, Ehlert S. & Mewes, J.-M. r2SCAN-3c: A “Swiss army knife” composite electronic-structure method. The Journal of Chemical Physics. 2021;154:064103.

    Not in our library10.1063/5.0040021

73. 73.

    Spicher S, Grimme S. Single-point Hessian calculations for improved vibrational frequencies and rigid-rotor-harmonic-oscillator thermodynamics. Journal of Chemical Theory and Computation. 2021;17:1701-1714.

    Not in our library10.1021/acs.jctc.0c01306

74. 74.

    Klamt A. Conductor-like screening model for real solvents: a new approach to the quantitative calculation of solvation phenomena. The Journal of Physical Chemistry. 1995;99:2224-2235.

    Not in our library10.1021/j100007a062

75. 75.

    Klamt A, Jonas V, Bürger T, Lohrenz JC. Refinement and parametrization of COSMO-RS. The Journal of Physical Chemistry A. 1998;102:5074-5085.

    Not in our library10.1021/jp980017s

76. 76.

    Barone V, Cossi M. Quantum calculation of molecular energies and energy gradients in solution by a conductor solvent model. The Journal of Physical Chemistry A. 1998;102:1995-2001.

    Not in our library10.1021/jp9716997

77. 77.

    Grimme S. Supramolecular binding thermodynamics by dispersion-corrected density functional theory. Chemistry–A European Journal. 2012;18:9955-9964.

    Not in our library10.1002/chem.201200497

78. 78.

    Open Source Data. https://www.aicures.mit.edu/data. Accessed: 2020-05-22 (2020).

    Not in our library

79. 79.

    Main protease structure and XChem fragment screen. https://www.diamond.ac.uk/covid-19/for-scientists/Main-protease-structure-and-XChem.html. Accessed: 2020-05-22.

    Not in our library

80. 80.

    Tokars, V. & Mesecar, A. QFRET-based primary biochemical high throughput screening assay to identify inhibitors of the SARS coronavirus 3C-like Protease (3CLPro). https://pubchem.ncbi.nlm.nih.gov/bioassay/1706.

    Not in our library

81. 81.

    Zampieri M, Zimmermann M, Claassen M, Sauer U. Nontargeted metabolomics reveals the multilevel response to antibiotic perturbations. Cell reports. 2017;19:1214-1228.

    Not in our library10.1016/j.celrep.2017.04.002

82. 82.

    Halgren, T. A. Merck molecular force field. I. Basis, form, scope, parameterization, and performance of MMFF94. Journal of computational chemistry17, 490–519, 10.1002/(SICI)1096-987X(199604)17:5/6<490::AID-JCC1>3.0.CO;2-P (1996).

    Not in our library

83. 83.

    Neese F. The ORCA program system. Wiley Interdisciplinary Reviews: Computational Molecular Science. 2012;2:73-78.

    Not in our library10.1002/wcms.81

84. 84.

    Neese F. Software update: the ORCA program system, version 4.0. Wiley Interdisciplinary Reviews: Computational Molecular Science. 2018;8:e1327.

    Not in our library10.1002/wcms.1327

85. 85.

    Kim Y, Kim WY. Universal structure conversion method for organic molecules: from atomic connectivity to three-dimensional geometry. Bulletin of the Korean Chemical Society. 2015;36:1769-1777.

    Not in our library10.1002/bkcs.10334

86. 86.

    Ehlert S, Stahn M, Spicher S, Grimme S. A robust and efficient implicit solvation model for fast semiempirical methods. Journal of Chemical Theory and Computation. 2021;17:4250-4261.

    Not in our library10.1021/acs.jctc.1c00471

87. 87.

    Neese F, Wennmohs F, Becker U, Riplinger C. The ORCA quantum chemistry program package. The Journal of Chemical Physics. 2020;152:224108.

    Not in our library10.1063/5.0004608

88. 88.

    Xu, M., Luo, S., Bengio, Y., Peng, J. & Tang, J. Learning neural generative dynamics for molecular conformation generation. In International Conference on Learning Representationshttps://openreview.net/forum?id=pAbm1qfheGk (2021).

    Not in our library

89. 89.

    Frederick KK, Marlow MS, Valentine KG, Wand AJ. Conformational entropy in molecular recognition by proteins. Nature. 2007;448:325-329.

    Not in our library10.1038/nature05959

90. 90.

    Murphy, K. P. Machine learning: a probabilistic perspective (MIT press, 2012).

    Not in our library

91. 91.

    Breiman L. Random forests. Machine learning. 2001;45:5-32.

    Not in our library10.1023/a:1010933404324

92. 92.

    Rogers D, Hahn M. Extended-connectivity fingerprints. Journal of chemical information and modeling. 2010;50:742-754.

    Not in our library10.1021/ci100050t

93. 93.

    Gilmer, J., Schoenholz, S. S., Riley, P. F., Vinyals, O. & Dahl, G. E. Neural message passing for quantum chemistry. In International Conference on Machine Learning, 70, 1263–1272, https://proceedings.mlr.press/v70/gilmer17a.html (PMLR, 2017)

    Not in our library

94. 94.

    Schütt KT, Sauceda HE, Kindermans P-J, Tkatchenko A, Müller K-R. SchNet–A deep learning architecture for molecules and materials. The Journal of Chemical Physics. 2018;148:241722.

    Not in our library10.1063/1.5019779

95. 95.

    Schütt, K. et al. SchNet: A continuous-filter convolutional neural network for modeling quantum interactions. In Advances in neural information processing systems, 991–1001, https://proceedings.neurips.cc/paper/2017/file/303ed4c69846ab36c2904d3ba8573050-Paper.pdf (2017).

    Not in our library

96. 96.

    Harvard Dataverse. 2021.

    Not in our library10.7910/dvn/n4vlql

97. 97.

    Harvard Dataverse. 2021.

    Not in our library10.7910/dvn/jngtdf

98. 98.

    Rappé AK, Casewit CJ, Colwell K, Goddard WA, Skiff WM. UFF, a full periodic table force field for molecular mechanics and molecular dynamics simulations. Journal of the American chemical society. 1992;114:10024-10035.

    Not in our library10.1021/ja00051a040

99. 99.

    Wang J, Wolf RM, Caldwell JW, Kollman PA, Case DA. Development and testing of a general amber force field. Journal of computational chemistry. 2004;25:1157-1174.

    Not in our library10.1002/jcc.20035

100. 100.
     Optimization of parameters for semiempirical methods VI: more modifications to the NDDO approximations and re-optimization of parameters · 2012

     No notice
101. 101.

     Wenlock, M. & Tomkinson, N. Experimental in vitro DMPK and physicochemical data on a set of publicly disclosed compounds. 10.6019/CHEMBL3301361.

     Not in our library

102. 102.

     Martins IF, Teixeira AL, Pinheiro L, Falcao AO. A Bayesian approach to in silico blood-brain barrier penetration modeling. Journal of chemical information and modeling. 2012;52:1686-1697.

     Not in our library10.1021/ci300124c

103. 103.

     Tox21 challenge. http://tripod.nih.gov/tox21/challenge/. Accessed 2017-09-27.

     Not in our library

104. 104.

     Richard AM, et al. ToxCast chemical landscape: paving the road to 21st century toxicology. Chemical research in toxicology. 2016;29:1225-1251.

     Not in our library10.1021/acs.chemrestox.6b00135

105. 105.
     The SIDER database of drugs and side effects · 2015

     No notice
106. 106.

     Novick PA, Ortiz OF, Poelman J, Abdulhay AY, Pande VS. SWEETLEAD: an in silico database of approved drugs, regulated chemicals, and herbal isolates for computer-aided drug discovery. PloS one. 2013;8:e79568.

     Not in our library10.1371/journal.pone.0079568

107. 107.

     Aggregate Analysis of ClincalTrials.gov (AACT) Database. https://aact.ctti-clinicaltrials.org/. Accessed 2017-09-27.

     Not in our library

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.