Abstract
Predicting the values of the potential energy surface (PES) for a given chemical system is essential to running the associated dynamics and modeling its evolution in time. To the purpose of modeling chemical reactions involving few atoms, this task is usually accomplished by fitting or interpolating a set of energies computed at different nuclear geometries through accurate, though computationally demanding, quantum-chemical calculations. Among the several approaches for choosing an appropriate set of geometries and energies, a new scheme has been recently proposed (Rampino S, J Phys Chem A 120:4683–4692, 2016) which is based on a regular sampling in a space-reduced bond-order (SRBO) domain rather than in the more conventional bond-length (BL) domain. In this work we address the performances of four machine-learning (ML) models, as opposed to pure mathematical fitting or interpolation schemes, in predicting the PES of a three-atom system modeling an atom-diatom exchange reaction when coupled to the SRBO sampling scheme. The models (two ensemble-learning, an automated ML, and a deep-learning one), trained on both SRBO and BL datasets, are shown to perform better than popular fitting or interpolation schemes and to give the best results if coupled to SRBO data.
The research leading to these results has received funding from Scuola Normale Superiore through project ‘DIVE: Development of Immersive approaches for the analysis of chemical bonding through Virtual-reality Environments’ (SNS18_B_RAMPINO) and program ‘Finanziamento a supporto della ricerca di base’ (SNS_RB_RAMPINO).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Jaquet, R.: Interpolation and fitting of potential energy surfaces: concepts, recipes and applications. In: Sax, A.F. (ed.) Potential Energy Surfaces. Lecture Notes in Chemistry, vol. 71, pp. 97–175. Springer, Berlin (1999). https://doi.org/10.1007/978-3-642-46879-7_3
Behler, J.: Neural network potential-energy surfaces for atomistic simulations. In: Chemical Modelling: Applications and Theory, vol. 7, pp. 1–41. The Royal Society of Chemistry, Cambridge (2010)
Handley, C.M., Popelier, P.L.A.: Potential energy surfaces fitted by artificial neural networks. The J. Phys. Chem. A 114(10), 3371–3383 (2010)
Hughes, Z.E., Thacker, J.C.R., Wilson, A.L., Popelier, P.L.A.: Description of potential energy surfaces of molecules using FFLUX machine learning models. J. Chem. Theor. Comput. 15(1), 116–126 (2019)
Raff, L., Komanduri, R., Hagan, M., Bukkapatnam, S.: Neural Networks in Chemical Reaction Dynamics. Oxford University Press, New York (2012)
Laganà, A., Costantini, A., Gervasi, O., Faginas Lago, N., Manuali, C., Rampino, S.: COMPCHEM: progress towards GEMS a grid empowered molecular simulator and beyond. J. Grid Comput. 8(4), 571–586 (2010)
Rampino, S.: Workflows and data models for atom diatom quantum reactive scattering calculations on the Grid. Ph.D. thesis, Università degli Studi di Perugia (2011)
Manuali, C., Laganà, A., Rampino, S.: GriF: a grid framework for a web service approach to reactive scattering. Comput. Phys. Commun. 181(7), 1179–1185 (2010)
Rampino, S., Faginas Lago, N., Laganà, A., Huarte-Larrañaga, F.: An extension of the grid empowered molecular simulator to quantum reactive scattering. J. Comput. Chem. 33(6), 708–714 (2012)
Rampino, S., Storchi, L., Laganà, A.: Automated simulation of gas-phase reactions on distributed and cloud computing infrastructures. In: Gervasi, O., et al. (eds.) ICCSA 2017. LNCS, vol. 10406, pp. 60–73. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-62398-6_5
Rampino, S., Skouteris, D., Laganà, A., Garcia, E.: A comparison of the isotope effect for the N + N\(_2\) reaction calculated on two potential energy surfaces. In: Gervasi, O., Murgante, B., Laganà, A., Taniar, D., Mun, Y., Gavrilova, M.L. (eds.) Computational Science and Its Applications - ICCSA 2008. Lecture Notes in Computer Science, vol. 5072, pp. 1081–1093. Springer, Berlin Heidelberg (2008)
Laganà, A., Faginas Lago, N., Rampino, S., Huarte Larrañaga, F., García, E.: Thermal rate coefficients in collinear versus bent transition state reactions: the N+N\(_2\) case study. Physica Scripta 78(5), 058116 (2008)
Rampino, S., Pirani, F., Garcia, E., Laganà, A.: A study of the impact of long range interactions on the reactivity of N + N\(_2\) using the Grid Empowered Molecular Simulator GEMS. Int. J. Web Grid Serv. 6(2), 196–212 (2010)
Laganà, A., Rampino, S.: A grid empowered virtual versus real experiment for the barrierless Li + FH \(\rightarrow \) LiF + H reaction. In: Murgante, B., et al. (eds.) ICCSA 2014. LNCS, vol. 8579, pp. 571–584. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-09144-0_39
Rossi, E., et al.: Code interoperability and standard data formats in quantum chemistry and quantum dynamics: the Q5/D5Cost data model. J. Comput. Chem. 35(8), 611–621 (2014)
Rampino, S., Monari, A., Rossi, E., Evangelisti, S., Laganà, A.: A priori modeling of chemical reactions on computational grid platforms: workflows and data models. Chem. Phys. 398, 192–198 (2012)
EGI: The European grid infrastructure. http://www.egi.eu/. Accessed 27 Feb 2019
Rampino, S., Skouteris, D., Laganà, A., García, E., Saracibar, A.: A comparison of the quantum state-specific efficiency of N + N\(_2\) reaction computed on different potential energy surfaces. Phys. Chem. Chem. Phys. 11, 1752–1757 (2009)
Rampino, S., Garcia, E., Pirani, F., Laganà, A.: Accurate quantum dynamics on grid platforms: some effects of long range interactions on the reactivity of N + N\(_2\). In: Taniar, D., Gervasi, O., Murgante, B., Pardede, E., Apduhan, B.O. (eds.) Computational Science and Its Applications - ICCSA 2010. Lecture Notes in Computer Science, vol. 6019, pp. 1–12. Springer, Berlin Heidelberg (2010). https://doi.org/10.1007/978-3-642-12189-0_1
Rampino, S., Skouteris, D., Laganà, A.: The O + O\(_2\) reaction: quantum detailed probabilities and thermal rate coefficients. Theor. Chem. Acc.: Theor. Comput. Model. 123(3/4), 249–256 (2009)
Rampino, S., Skouteris, D., Laganà, A.: Microscopic branching processes: the O + O\(_2\) reaction and its relaxed potential representations. Int. J. Quantum Chem. 110(2), 358–367 (2010)
Rampino, S., Pastore, M., Garcia, E., Pacifici, L., Laganà, A.: On the temperature dependence of the rate coefficient of formation of C\(_2^+\) from C + CH\(^+\). Monthly Not. Roy. Astron. Soc. 460(3), 2368–2375 (2016)
Pacifici, L., Pastore, M., Garcia, E., Laganà, A., Rampino, S.: A dynamics investigation of the C + CH\(^+\)\(\rightarrow \) C\(_2^+\) + H reaction on an ab initio bond-order like potential. J. Phys. Chem. A 120(27), 5125–5135 (2016)
Rampino, S., Suleimanov, Y.V.: Thermal rate coefficients for the astrochemical process C + CH\(^+\)\(\rightarrow \) C\(_2^+\) + H by ring polymer molecular dynamics. J. Phys. Chem. A 120(50), 9887–9893 (2016)
Rampino, S.: Configuration-space sampling in potential energy surface fitting: a space-reduced bond-order grid approach. J. Phys. Chem. A 120(27), 4683–4692 (2016)
Pauling, L.: Atomic radii and interatomic distances in metals. J. Am. Chem. Soc. 69(3), 542–553 (1947)
Garcia, E., Laganà, A.: Diatomic potential functions for triatomic scattering. Mol. Phys. 56(3), 621–627 (1985)
Garcia, E., Laganà, A.: A new bond-order functional form for triatomic molecules. Mol. Phys. 56(3), 629–639 (1985)
Laganà, A.: A rotating bond order formulation of the atom diatom potential energy surface. J. Chem. Phys. 95(3), 2216–2217 (1991)
Laganà, A., Ochoa de Aspuru, G., Garcia, E.: The largest angle generalization of the rotating bond order potential: three different atom reactions. J. Chem. Phys. 108(10), 3886–3896 (1998)
Laganà, A., Crocchianti, S., Faginas Lago, N., Pacifici, L., Ferraro, G.: A nonorthogonal coordinate approach to atom-diatom parallel reactive scattering calculations. Collect. Czechoslovak Chem. Commun. 68(2), 307–330 (2003)
Rampino, S., Laganà, A.: Bond order uniform grids for quantum reactive scattering. Int. J. Quantum Chem. 112(7), 1818–1828 (2012)
Pedregosa, F., et al.: Scikit-learn: machine learning in Python. J. Mach. Learn. Res. 12, 2825–2830 (2011)
Feurer, M., Klein, A., Eggensperger, K., Springenberg, J., Blum, M., Hutter, F.: Efficient and robust automated machine learning. In: Cortes, C., Lawrence, N.D., Lee, D.D., Sugiyama, M., Garnett, R. (eds.) Advances in Neural Information Processing Systems, vol. 28, pp. 2962–2970. Curran Associates, Inc. (2015)
Chollet, F.: Keras (2015). https://github.com/fchollet/keras. Accessed 27 Feb 2019
Aguado, A., Paniagua, M.: A new functional form to obtain analytical potentials of triatomic molecules. J. Chem. Phys. 96(2), 1265–1275 (1992)
Aguado, A., Tablero, C., Paniagua, M.: Global fit of ab initio potential energy surfaces I. Triatomic systems. Comput. Phys. Commun. 108(2–3), 259–266 (1998)
Shepard, D.: A two-dimensional interpolation function for irregularly-spaced data. In: Proceedings of the 1968 23rd ACM National Conference, ACM 1968, pp. 517–524. ACM, New York (1968)
Surowiecki, J.: The Wisdom of Crowds: Why the Many Are Smarter Than the Few and How Collective Wisdom Shapes Business, Economies, Societies, and Nations. Doubleday & Co, New York (2004)
Breiman, L.: Random forests. Mach. Learn. 45(1), 5–32 (2001)
Quinlan, J.R.: Induction of decision trees. Mach. Learn. 1(1), 81–106 (1986)
Chen, T., Guestrin, C.: XGBoost: a scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2016, pp. 785–794. ACM, New York (2016)
Snoek, J., Larochelle, H., Adams, R.P.: Practical Bayesian optimization of machine learning algorithms. In: Proceedings of the 25th International Conference on Neural Information Processing Systems, NIPS 2012, vol. 2, pp. 2951–2959. Curran Associates Inc., USA (2012)
Brochu, E., Cora, V.M., de Freitas, N.: A tutorial on Bayesian optimization of expensive cost functions, with application to active user modeling and hierarchical reinforcement learning. eprint arXiv:1012.2599, arXiv.org, December 2010
Feurer, M., Eggensperger, K., Falkner, S., Lindauer, M., Hutter, F.: Practical automated machine learning for the AutoML challenge 2018. In: ICML 2018 AutoML Workshop (2018)
Esteva, A., et al.: A guide to deep learning in healthcare. Nat. Med. 25(1), 24–29 (2019)
Zou, J., Huss, M., Abid, A., Mohammadi, P., Torkamani, A., Telenti, A.: A primer on deep learning in genomics. Nat. Genet. 51(1), 12–18 (2019)
Nash, W., Drummond, T., Birbilis, N.: A review of deep learning in the study of materials degradation. npj Mater. Degrad. 2(1), 37 (2018)
LeCun, Y., Bengio, Y., Hinton, G.: Deep learning. Nature 521(7553), 436–444 (2015)
Salvadori, A., Fusè, M., Mancini, G., Rampino, S., Barone, V.: Diving into chemical bonding: an immersive analysis of the electron charge rearrangement through virtual reality. J. Comput. Chem. 39(31), 2607–2617 (2018)
Salvadori, A., et al.: A walk through chemistry: exploring potential-energy surfaces with virtual reality (2019, in preparation)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Licari, D., Rampino, S., Barone, V. (2019). Machine Learning of Potential-Energy Surfaces Within a Bond-Order Sampling Scheme. In: Misra, S., et al. Computational Science and Its Applications – ICCSA 2019. ICCSA 2019. Lecture Notes in Computer Science(), vol 11624. Springer, Cham. https://doi.org/10.1007/978-3-030-24311-1_28
Download citation
DOI: https://doi.org/10.1007/978-3-030-24311-1_28
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-24310-4
Online ISBN: 978-3-030-24311-1
eBook Packages: Computer ScienceComputer Science (R0)