journal article·2024-04-25Open access

Tell Machine Learning Potentials What They Are Needed For: Simulation-Oriented Training Exemplified for Glycine

Fuchun Ge, Ran Wang, Chen Qu, Peikun Zheng, Apurba Nandi, Riccardo Conte, Paul L. Houston, Joel M. Bowman, Pavlo O. Dral

Abstract

Machine learning potentials (MLPs) are widely applied as an efficient alternative way to represent potential energy surfaces (PESs) in many chemical simulations. The MLPs are often evaluated with the root-mean-square errors on the test set drawn from the same distribution as the training data. Here, we systematically investigate the relationship between such test errors and the simulation accuracy with MLPs on an example of a full-dimensional, global PES for the glycine amino acid. Our results show that the errors in the test set do not unambiguously reflect the MLP performance in different simulation tasks, such as relative conformer energies, barriers, vibrational levels, and zero-point vibrational energies. We also offer an easily accessible solution for improving the MLP quality in a simulation-oriented manner, yielding the most precise relative conformer energies and barriers. This solution also passed the stringent test by diffusion Monte Carlo simulations.
Source: openalex_inverted_index · Quality: baseline

Research context

OpenAlex topics: Machine Learning in Materials Science · Advanced Chemical Physics Studies · Spectroscopy and Quantum Chemical Studies

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

Affiliations

State Key Laboratory of Physical Chemistry of Solid Surfaces, College of Chemistry and Chemical Engineering, Fujian Provincial Key Laboratory of Theoretical and Computational Chemistry, and Innovation Laboratory for Sciences and Technologies of Energy Materials of Fujian Province (IKKEM), Xiamen University, Xiamen, Fujian 361005, China

Independent Researcher, Toronto, Ontario M9B0E3, Canada

Department of Chemistry and Cherry L. Emerson Center for Scientific Computation, Emory University, Atlanta, Georgia 30322, United States

Department of Physics and Materials Science, University of Luxembourg, Luxembourg City L-1511, Luxembourg

Dipartimento di Chimica, Università degli Studi di Milano, via Golgi 19, 20133 Milano, Italy

Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York 14853, United States

Funding

Universit? degli Studi di Milano · National Aeronautics and Space Administration · Fundamental Research Funds for the Central Universities · National Natural Science Foundation of China · Innovation Laboratory for Sciences and Technologies of Energy Materials of Fujian Province

References (85)

Show source-provided references
1. Born M. · 1954
2. 10.1002/andp.1927389200210.1002/andp.19273892002 ↗
3. Murrell J. N. · 1985
4. 10.1080/0144235X.2016.119510210.1080/0144235X.2016.1195102 ↗
5. 10.1002/9780470142943.ch110.1002/9780470142943.ch1 ↗
6. 10.1088/0031-9112/36/12/02910.1088/0031-9112/36/12/029 ↗
7. 10.1103/RevModPhys.61.66910.1103/RevModPhys.61.669 ↗
8. 10.1002/9781119518068.ch510.1002/9781119518068.ch5 ↗
9. 10.1063/1.46047110.1063/1.460471 ↗
10. 10.1063/1.48144210.1063/1.481442 ↗
11. 10.1007/s00214-013-1429-610.1007/s00214-013-1429-6 ↗
12. 10.1016/0009-2614(94)00939-210.1016/0009-2614(94)00939-2 ↗
13. 10.1080/0144235080204544610.1080/01442350802045446 ↗
14. 10.1063/1.243721410.1063/1.2437214 ↗
15. 10.1021/jp076716z10.1021/jp076716z ↗
16. 10.1021/acs.jctc.8b0057110.1021/acs.jctc.8b00571 ↗
17. 10.1007/s00214-002-0383-510.1007/s00214-002-0383-5 ↗
18. 10.1039/c0cp02722g10.1039/c0cp02722g ↗
19. 10.1021/jp406944810.1021/jp4069448 ↗
20. 10.1021/acs.jpca.5b0251010.1021/acs.jpca.5b02510 ↗
21. 10.1021/ja408723a10.1021/ja408723a ↗
22. 10.1063/1.499067310.1063/1.4990673 ↗
23. 10.1016/j.cplett.2020.13761010.1016/j.cplett.2020.137610 ↗
24. 10.1021/cr030718810.1021/cr0307188 ↗
25. 10.1039/C9SC05116C10.1039/C9SC05116C ↗
26. 10.1039/cs996250032110.1039/cs9962500321 ↗
27. 10.1016/j.sbi.2020.07.00310.1016/j.sbi.2020.07.003 ↗
28. 10.1021/acs.jctc.2c0114310.1021/acs.jctc.2c01143 ↗
29. 10.1021/ct600140b10.1021/ct600140b ↗
30. 10.1021/acs.jpclett.0c0098910.1021/acs.jpclett.0c00989 ↗
31. 10.1021/ja710290710.1021/ja7102907 ↗
32. 10.1021/jacs.1c0529310.1021/jacs.1c05293 ↗
33. 10.1038/s41467-020-19497-z10.1038/s41467-020-19497-z ↗
34. 10.1146/annurev-physchem-050317-02121910.1146/annurev-physchem-050317-021219 ↗
35. 10.1103/PhysRevLett.98.14640110.1103/PhysRevLett.98.146401 ↗
36. 10.1103/PhysRevB.87.18411510.1103/PhysRevB.87.184115 ↗
37. 10.1021/acs.chemrev.0c0111110.1021/acs.chemrev.0c01111 ↗
38. 10.1021/acs.chemrev.1c0002110.1021/acs.chemrev.1c00021 ↗
39. 10.1021/acs.chemrev.0c0066510.1021/acs.chemrev.0c00665 ↗
40. 10.1002/wcms.156410.1002/wcms.1564 ↗
41. 10.1080/0144235090323492310.1080/01442350903234923 ↗
42. 10.1146/annurev-physchem-050317-02113910.1146/annurev-physchem-050317-021139 ↗
43. 10.1038/s41563-020-0777-610.1038/s41563-020-0777-6 ↗
44. 10.1021/acs.chemrev.0c0086810.1021/acs.chemrev.0c00868 ↗
45. Ceriotti M. · 2018
46. 10.1039/D1SC03564A10.1039/D1SC03564A ↗
47. Zeng J. · 2023 · 27910.1016/B978-0-323-90049-2.00001-9 ↗
48. Hou Y.-F. · 2023 · 29510.1016/B978-0-323-90049-2.00020-2 ↗
49. Tallec G. · 2023 · 25310.1016/B978-0-323-90049-2.00021-4 ↗
50. 10.1021/ct900491710.1021/ct9004917 ↗
51. 10.1063/5.013444210.1063/5.0134442 ↗
52. 10.1088/2632-2153/abc9fe10.1088/2632-2153/abc9fe ↗
53. 10.1039/C6SC05720A10.1039/C6SC05720A ↗
54. 10.1103/PhysRevLett.104.13640310.1103/PhysRevLett.104.136403 ↗
55. 10.1063/1.498953610.1063/1.4989536 ↗
56. 10.1126/sciadv.160301510.1126/sciadv.1603015 ↗
57. 10.1016/j.cpc.2019.02.00710.1016/j.cpc.2019.02.007 ↗
58. 10.1021/acs.jctc.2c0103810.1021/acs.jctc.2c01038 ↗
59. 10.1021/acs.jpca.9b0872310.1021/acs.jpca.9b08723 ↗
60. 10.1063/5.003851610.1063/5.0038516 ↗
61. 10.1021/acs.jctc.1c0064710.1021/acs.jctc.1c00647 ↗
62. 10.1038/s41467-021-25342-810.1038/s41467-021-25342-8 ↗
63. 10.1088/2632-2153/abfd9610.1088/2632-2153/abfd96 ↗
64. 10.1039/D3CP03515H10.1039/D3CP03515H ↗
65. 10.1063/5.008920010.1063/5.0089200 ↗
66. 10.1063/5.000649810.1063/5.0006498 ↗
67. 10.1063/5.003717510.1063/5.0037175 ↗
68. 10.1039/C6SC05720A10.1039/C6SC05720A ↗
69. 10.1103/PhysRevLett.125.16600110.1103/PhysRevLett.125.166001 ↗
70. Batatia I. · 2022
71. 10.1002/jcc.2637510.1002/jcc.26375 ↗
72. 10.1021/acs.jctc.3c0120310.1021/acs.jctc.3c01203 ↗
73. 10.1021/acs.jcim.0c0045110.1021/acs.jcim.0c00451 ↗
74. 10.1021/acs.jctc.2c0109710.1021/acs.jctc.2c01097 ↗
75. 10.1063/5.008050610.1063/5.0080506 ↗
76. 10.1021/acs.jctc.4c0005410.1021/acs.jctc.4c00054 ↗
77. 10.1038/s41597-020-0473-z10.1038/s41597-020-0473-z ↗
78. 10.1063/1.502380210.1063/1.5023802 ↗
79. 10.1002/qua.2483610.1002/qua.24836 ↗
80. 10.1103/PhysRevLett.114.09640510.1103/PhysRevLett.114.096405 ↗
81. 10.1103/PhysRevMaterials.3.02380410.1103/PhysRevMaterials.3.023804 ↗
82. 10.1039/C7SC02267K10.1039/C7SC02267K ↗
83. 10.1038/s41467-019-10827-410.1038/s41467-019-10827-4 ↗
84. Paszke A. · 2019 · 32 · 8026
85. Frisch M. J. · 2016