Graph neural networks for molecular and materials representation

Abstract

Material molecular representation (MMR) plays an important role in material property or chemical reaction prediction. However, traditional expert-designed MMR methods face challenges in dealing with high dimensionality and heterogeneity of material data, leading to limited generalization capabilities and insufficient information representation. In recent years, graph neural networks (GNNs), a deep learning algorithm specifically designed for graph structures, have made inroads into the field of MMR. It will be instructive and inspiring to conduct a survey on various GNNs used for MMR. To achieve this objective, we compare GNNs with conventional MMR methods and illustrate the advantages of GNNs, such as their expressiveness and adaptability. In addition, we systematically classify and summarize the methods and applications of GNNs. Finally, we provide our insights into future research directions, taking into account the characteristics of molecular data and the inherent drawbacks of GNNs. This comprehensive survey is intended to present a holistic view of GNNs for MMR, focusing on the core concepts, the main techniques, and the future trends in this area.

Keywords

Material molecular representation / material property / reaction prediction / graph neural networks / expressiveness / adaptability

References

[1] Zhang TY.Informatics is fueling new materials discovery.J Mater Inf2021;1:6

[2] Hara K,Kurotani A,Kikuchi J.Materials informatics approach using domain modelling for exploring structure-property relationships of polymers.Sci Rep2022;12:10558 PMCID:PMC9217937

[3] Kuz’min V,Ognichenko L.Simplex representation of molecular structure as universal QSAR/QSPR tool.Struct Chem2021;32:1365-92 PMCID:PMC8218296

[4] Keyvanpour MR.An analysis of QSAR research based on machine learning concepts.Curr Drug Discov Technol2021;18:17-30

[5] Poulson BG,Sharfalddin A.Cyclodextrins: structural, chemical, and physical properties, and applications.Polysaccharides2022;3:1-31

[6] Gao H.Graph U-Nets.IEEE Trans Pattern Anal Mach Intell2022;44:4948-60

[7] Lee JB,Kong X.Graph classification using structural attention. Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining; London, United Kingdom; 2018. pp. 1666-74.

[8] Gilmer J,Riley PF,Dahl GE. Neural message passing for quantum chemistry. Proceedings of the 34th International Conference on Machine Learning; Sydney, NSW, Australia; 2017. pp. 1263-72. Available from: https://proceedings.mlr.press/v70/gilmer17a [Last accessed on 7 Jun 2023]

[9] Wei X,Jia Z.High-cycle fatigue S-N curve prediction of steels based on a transfer learning-guided convolutional neural network.J Mater Inf2022;2:9

[10] Luo H,Fang W,Zhang B.Convolutional neural networks: computer vision-based workforce activity assessment in construction.Autom Constr2018;94:282-9

[11] Zhang J,Wu X,Hwang J.Coarse-to-fine multiscale fusion network for single image deraining.J Electron Imag2022;31:043003

[12] Wu X,Li Q,Wang J.Face aging with pixel-level alignment GAN.Appl Intell2022;52:14665-78

[13] Khurana D,Khatter K.Natural language processing: state of the art, current trends and challenges.Multimed Tools Appl2023;82:3713-44 PMCID:PMC9281254

[14] Wu X,Wang J,Guo Y.MKD: Mixup-based knowledge distillation for mandarin end-to-end speech recognition.Algorithms2022;15:160

[15] Wu X,Zhao M,Guo Y.STR transformer: a cross-domain transformer for scene text recognition.Appl Intell2023;53:3444-58

[16] Zhou J,Hu S.Graph neural networks: a review of methods and applications.AI Open2020;1:57-81

[17] Wu Z,Chen F,Zhang C.A comprehensive survey on graph neural networks.IEEE Trans Neural Netw Learn Syst2021;32:4-24

[18] Zagidullin B,Guan Y,Tang J.Comparative analysis of molecular fingerprints in prediction of drug combination effects.Brief Bioinform2021;22:bbab291 PMCID:PMC8574997

[19] Melville JL,Hirst JD.Similarity by compression.J Chem Inf Model2007;47:25-33

[20] Duvenaud D,Aguilera-Iparraguirre J. Convolutional networks on graphs for learning molecular fingerprints. Proceedings of the 28th International Conference on Neural Information Processing Systems; Montreal, Canada; 2015. pp. 2224-32. Available from: https://dl.acm.org/doi/10.5555/2969442.2969488 [Last accessed on 8 Jun 2023]

[21] Ding Y,Guo C,Wang J.Molecular fingerprint-based machine learning assisted QSAR model development for prediction of ionic liquid properties.J Mol Liq2021;326:115212

[22] Maggiora G,Stumpfe D.Molecular similarity in medicinal chemistry.J Med Chem2014;57:3186-204

[23] Rogers D.Extended-connectivity fingerprints.J Chem Inf Model2010;50:742-54

[24] Rush TS 3rd,Mosyak L.A shape-based 3-D scaffold hopping method and its application to a bacterial protein-protein interaction.J Med Chem2005;48:1489-95

[25] Durant JL,Henry DR.Reoptimization of MDL keys for use in drug discovery.J Chem Inf Comput Sci2002;42:1273-80

[26] Liu X,Huang R.Long short-term memory recurrent neural network for pharmacokinetic-pharmacodynamic modeling.Int J Clin Pharmacol Ther2021;59:138-46

[27] Goulas A,Hilgetag CC.Bio-instantiated recurrent neural networks: Integrating neurobiology-based network topology in artificial networks.Neural Netw2021;142:608-18

[28] Weininger D.SMILES, a chemical language and information system. 1. introduction to methodology and encoding rules.J Chem Inf Comput Sci1988;28:31-6

[29] Heller SR,Pletnev I,Tchekhovskoi D.InChI, the IUPAC international chemical identifier.J Cheminform2015;7:23 PMCID:PMC4486400

[30] Lin X,Wang ZJ,Zeng X.A novel molecular representation with BiGRU neural networks for learning atom.Brief Bioinform2020;21:2099-111

[31] Feng YH.Prediction of drug-drug interaction using an attention-based graph neural network on drug molecular graphs.Molecules2022;27:3004 PMCID:PMC9105425

[32] Chuang KV,Keiser MJ.Learning molecular representations for medicinal chemistry.J Med Chem2020;63:8705-22

[33] Bengio Y,Vincent P.Representation learning: a review and new perspectives.IEEE Trans Pattern Anal Mach Intell2013;35:1798-828

[34] Zhang Z,Xu Y,Yang J.Marginal representation learning with graph structure self-adaptation.IEEE Trans Neural Netw Learn Syst2018;29:4645-59

[35] Xie Y,Gong M,Yu B.Multi-task network representation learning.Front Neurosci2020;14:1 PMCID:PMC6989613

[36] Wang S,Gong M.Multi-task learning based network embedding.Front Neurosci2019;13:1387 PMCID:PMC6971216

[37] Khasanova R. Graph-based isometry invariant representation learning. Proceedings of the 34th International Conference on Machine Learning; Sydney, NSW, Australia; 2017. pp. 1847-56. Available from: http://proceedings.mlr.press/v70/khasanova17a.html?ref=https://githubhelp.com [Last accessed on 8 Jun 2023]

[38] Lee S.Scale-invariant representation of machine learning.Phys Rev E2022;105:044306

[39] Batra R,Ramprasad R.Emerging materials intelligence ecosystems propelled by machine learning.Nat Rev Mater2021;6:655-78

[40] Agrawal A.Deep materials informatics: applications of deep learning in materials science.MRS Commun2019;9:779-92

[41] Chen C,Zuo Y,Ong SP.Graph networks as a universal machine learning framework for molecules and crystals.Chem Mater2019;31:3564-72

[42] Park CW.Developing an improved crystal graph convolutional neural network framework for accelerated materials discovery.Phys Rev Materials2020;4:063801

[43] Sumpter BG.Neural networks and graph theory as computational tools for predicting polymer properties.Macromol Theory Simul3:363-78

[44] Xie T.Crystal graph convolutional neural networks for an accurate and interpretable prediction of material properties.Phys Rev Lett2018;120:145301

[45] Coley CW,Rogers L.A graph-convolutional neural network model for the prediction of chemical reactivity.Chem Sci2019;10:370-7 PMCID:PMC6335848

[46] Wu Z,Feinberg EN.MoleculeNet: a benchmark for molecular machine learning.Chem Sci2018;9:513-30 PMCID:PMC5868307

[47] Schmidt J,Botti S.Recent advances and applications of machine learning in solid-state materials science.npj Comput Mater2019;5:83

[48] Defferrard M,Vandergheynst P. Convolutional neural networks on graphs with fast localized spectral filtering. Available from: https://arxiv.org/abs/1606.09375 [Last accessed on 8 Jun 2023]

[49] Hammond DK,Gribonval R.Wavelets on graphs via spectral graph theory.Appl Comput Harmon A2011;30:129-50

[50] Li R,Zhu F.Adaptive graph convolutional neural networks.AAAI2018;32

[51] Monti F,Masci J,Svoboda J. Bronstein MM. Geometric deep learning on graphs and manifolds using mixture model CNNs. Available from: https://ieeexplore.ieee.org/document/8100059 [Last accessed on 8 Jun 2023]

[52] Atwood J. Diffusion-convolutional neural networks. Available from: https://proceedings.neurips.cc/paper_files/paper/2016/hash/390e982518a50e280d8e2b535462ec1f-Abstract.html [Last accessed on 8 Jun 2023]

[53] Vaswani A,Parmar N. Attention is all you need. Available from: https://proceedings.neurips.cc/paper_files/paper/2017/hash/3f5ee243547dee91fbd053c1c4a845aa-Abstract.html [Last accessed on 8 Jun 2023]

[54] Veličković P,Casanova A,Liò P. Graph attention networks. Available from: https://arxiv.org/abs/1710.10903 [Last accessed on 8 Jun 2023]

[55] Lusci A,Baldi P.Deep architectures and deep learning in chemoinformatics: the prediction of aqueous solubility for drug-like molecules.J Chem Inf Model2013;53:1563-75 PMCID:PMC3739985

[56] Altae-Tran H,Pappu AS.Low data drug discovery with one-shot learning.ACS Cent Sci2017;3:283-93 PMCID:PMC5408335

[57] Segler MHS,Tyrchan C.Generating focused molecule libraries for drug discovery with recurrent neural networks.ACS Cent Sci2018;4:120-31 PMCID:PMC5785775

[58] Rahimi A,Baldwin T.Semi-supervised user geolocation via graph convolutional networks. Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers); Melbourne, Australia; 2018. pp. 2009-19.

[59] Li G,Thabet A. DeepGCNs: can GCNs go as deep as CNNs? 2019 IEEE/CVF International Conference on Computer Vision (ICCV); Seoul, Korea (South); 2019. pp. 9266-76. Available from: https://openaccess.thecvf.com/content_ICCV_2019/html/Li_DeepGCNs_Can_GCNs_Go_As_Deep_As_CNNs_ICCV_2019_paper.html [Last accessed on 8 Jun 2023]

[60] Morris C,Fey M.Weisfeiler and leman go neural: higher-order graph neural networks.AAAI2019;33:4602-9

[61] Papp PA,Faber L. DropGNN: random dropouts increase the expressiveness of graph neural networks. Available from: https://proceedings.neurips.cc/paper/2021/hash/b8b2926bd27d4307569ad119b6025f94-Abstract.html [Last accessed on 8 Jun 2023]

[62] Rong Y,Xu T. DropEdge: towards deep graph convolutional networks on node classification. Available from: https://openreview.net/forum?id=Hkx1qkrKPr [Last accessed on 8 Jun 2023]

[63] Sun Q,Peng H. SUGAR: subgraph neural network with reinforcement pooling and self-supervised mutual information mechanism. Available from: https://arxiv.org/abs/2101.08170 [Last accessed on 8 Jun 2023]

[64] Bevilacqua B,Lim D. Equivariant subgraph aggregation networks. Available from: https://openreview.net/forum?id=dFbKQaRk15w [Last accessed on 8 Jun 2023]

[65] Gasteiger J,Günnemann S. Directional message passing on molecular graphs via synthetic coordinates. Available from: https://proceedings.neurips.cc/paper/2021/hash/82489c9737cc245530c7a6ebef3753ec-Abstract.html [Last accessed on 8 Jun 2023]

[66] Liu Y,Liu M. Spherical message passing for 3D molecular graphs. Available from: https://par.nsf.gov/servlets/purl/10353844 [Last accessed on 8 Jun 2023]

[67] Gasteiger J,Günnemann S. Gemnet: universal directional graph neural networks for molecules. Available from: https://proceedings.neurips.cc/paper/2021/hash/35cf8659cfcb13224cbd47863a34fc58-Abstract.html [Last accessed on 8 Jun 2023]

[68] Vasudevan RK,Mehta A.Materials science in the AI age: high-throughput library generation, machine learning and a pathway from correlations to the underpinning physics.MRS Commun2019;9:10.1557/mrc.2019.95 PMCID:PMC7067066

[69] Lu S,Ouyang Y,Li Q.Accelerated discovery of stable lead-free hybrid organic-inorganic perovskites via machine learning.Nat Commun2018;9:3405 PMCID:PMC6109147

[70] Xie T,Wang Y.Accelerating amorphous polymer electrolyte screening by learning to reduce errors in molecular dynamics simulated properties.Nat Commun2022;13:3415 PMCID:PMC9197847

[71] Xie T.Hierarchical visualization of materials space with graph convolutional neural networks.J Chem Phys2018;149:174111

[72] Gómez-Bombarelli R,Duvenaud D.Automatic chemical design using a data-driven continuous representation of molecules.ACS Cent Sci2018;4:268-76 PMCID:PMC5833007

[73] Zhao H,Huang H.A robotic platform for the synthesis of colloidal nanocrystals.Nat Synth2023;2:505-14

[74] Lee YJ,Kim SB.Generative adversarial networks for de novo molecular design.Mol Inform2021;40:e2100045

[75] Patel RA,Webb MA.Featurization strategies for polymer sequence or composition design by machine learning.Mol Syst Des Eng2022;7:661-76

[76] Putin E,Ivanenkov Y.Reinforced adversarial neural computer for de novo molecular design.J Chem Inf Model2018;58:1194-204

[77] Zhao Y,Hu M.High-throughput discovery of novel cubic crystal materials using deep generative neural networks.Adv Sci2021;8:e2100566 PMCID:PMC8529451

[78] You J,Ren X,Leskovec J. GraphRNN: generating realistic graphs with deep auto-regressive models. Proceedings of the 35th International Conference on Machine Learning; 2018. pp. 5708-17. Available from: http://proceedings.mlr.press/v80/you18a.html?ref=https://githubhelp.com [Last accessed on 8 Jun 2023]

[79] Lai X,Wang K,Yu D.MGRNN: structure generation of molecules based on graph recurrent neural networks.Mol Inform2021;40:e2100091

[80] Sanchez-Lengeling B.Inverse molecular design using machine learning: generative models for matter engineering.Science2018;361:360-5

[81] Lin X,Yang Y.Molecular distance matrix prediction based on graph convolutional networks.J Mol Struct2022;1257:132540

[82] Gong S,Tian Y,Liu G.Rapid enthalpy prediction of transition states using molecular graph convolutional network.AIChE Journal2023;69

[83] Jain A,Hautier G.Commentary: the materials project: a materials genome approach to accelerating materials innovation.APL Materials2013;1:011002

[84] Im S,Kim W.Neural network constitutive model for crystal structures.Comput Mech2021;67:185-206

[85] Dunn A,Ganose A,Jain A.Benchmarking materials property prediction methods: the matbench test set and automatminer reference algorithm.npj Comput Mater2020;6:138

[86] Fung V,Juarez E.Benchmarking graph neural networks for materials chemistry.npj Comput Mater2021;7:84

[87] Choudhary K.Atomistic line graph neural network for improved materials property predictions.npj Comput Mater2021;7:185

[88] Louis SY,Nasiri A.Graph convolutional neural networks with global attention for improved materials property prediction.Phys Chem Chem Phys2020;22:18141-8

[89] Li Y,Yang X.Introducing block design in graph neural networks for molecular properties prediction.Chem Eng J2021;414:128817

[90] Trieu HL,Ananiadou S.BioVAE: a pre-trained latent variable language model for biomedical text mining.Bioinformatics2022;38:872-4 PMCID:PMC8756089

[91] Zhang ZC,Zhou T.Pre-trained language model augmented adversarial training network for Chinese clinical event detection.Math Biosci Eng2020;17:2825-41

[92] Lee J,Kim S.BioBERT: a pre-trained biomedical language representation model for biomedical text mining.Bioinformatics2020;36:1234-40 PMCID:PMC7703786

[93] Wang Y,Cao Z.Molecular contrastive learning of representations via graph neural networks.Nat Mach Intell2022;4:279-87

[94] Ding Z,Ma T,Ma W.Predicting the hydrogen release ability of LiBH4 -based mixtures by ensemble machine learning.Energy Stor Mater2020;27:466-77

[95] Blum LC.970 million druglike small molecules for virtual screening in the chemical universe database GDB-13.J Am Chem Soc2009;131:8732-3

[96] Ruddigkeit L,Blum LC.Enumeration of 166 billion organic small molecules in the chemical universe database GDB-17.J Chem Inf Model2012;52:2864-75

[97] Ramakrishnan R,Rupp M.Quantum chemistry structures and properties of 134 kilo molecules.Sci Data2014;1:140022 PMCID:PMC4322582

[98] Gan Y,Sun Z.Prediction of the atomic structure and thermoelectric performance for semiconducting Ge¹Sb⁶Te¹⁰ from DFT calculations.J Mater Inf2021;1:2

[99] Elegbeleye IF,Maphanga RR.Density functional theory study of optical and electronic properties of (TiO₂)_n=5,8,68 clusters for application in solar cells.Molecules2021;26:955 PMCID:PMC7916945

[100] Liao R,Urtasun R. LanczosNet: multi-scale deep graph convolutional networks. Available from: https://openreview.net/forum?id=BkedznAqKQ [Last accessed on 8 Jun 2023]

[101] Louis SY,Joshi RP,Kumar N.Accurate prediction of voltage of battery electrode materials using attention-based graph neural networks.ACS Appl Mater Interfaces2022;14:26587-94

[102] Omee SS,Fu N.Scalable deeper graph neural networks for high-performance materials property prediction.Patterns2022;3:100491 PMCID:PMC9122959

[103] Breuck P, Heymans G, Rignanese G. Accurate experimental band gap predictions with multifidelity correction learning.J Mater Inf2022;2:10

[104] Shang C,Tong Q,Song M.Multi-view spectral graph convolution with consistent edge attention for molecular modeling.Neurocomputing2021;445:12-25

[105] Tang B,Fang M,Wu Z.A self-attention based message passing neural network for predicting molecular lipophilicity and aqueous solubility.J Cheminform2020;12:15 PMCID:PMC7035778

[106] Huuskonen J.Estimation of aqueous solubility for a diverse set of organic compounds based on molecular topology.J Chem Inf Comput Sci2000;40:773-7

[107] Micheli A.Neural network for graphs: a contextual constructive approach.IEEE Trans Neural Netw2009;20:498-511

[108] Coley CW,Green WH,Jensen KF.Convolutional embedding of attributed molecular graphs for physical property prediction.J Chem Inf Model2017;57:1757-72

[109] Meng M,Li Z,Bian Y.Property prediction of molecules in graph convolutional neural network expansion. 2019 IEEE 10th International Conference on Software Engineering and Service Science (ICSESS); Beijing, China.

[110] Dai M,Liang Y.Graph neural networks for an accurate and interpretable prediction of the properties of polycrystalline materials.npj Comput Mater2021;7:103

[111] Groom CR,Lightfoot MP.The cambridge structural database.Acta Crystallogr B Struct Sci Cryst Eng Mater2016;72:171-9 PMCID:PMC4822653

[112] Gražulis S,Merkys A.Crystallography open database (COD): an open-access collection of crystal structures and platform for world-wide collaboration.Nucleic Acids Res2012;40:D420-7 PMCID:PMC3245043

[113] Hao Z,Huang Z.ASGN: An active semi-supervised graph neural network for molecular property prediction. Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining; Virtual Event, CA, USA; 2020. pp. 731-52.

[114] Park CW,Vandermause J,Kozinsky B.Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture.npj Comput Mater2021;7:73

[115] Vugmeyster Y,Xu X.Absorption, distribution, metabolism, and excretion (ADME) studies of biotherapeutics for autoimmune and inflammatory conditions.AAPS J2012;14:714-27 PMCID:PMC3475869

[116] Delaney JS.ESOL: estimating aqueous solubility directly from molecular structure.J Chem Inf Comput Sci2004;44:1000-5

[117] Bergström CA,Lazorova L,Luthman K.Absorption classification of oral drugs based on molecular surface properties.J Med Chem2003;46:558-70

[118] Otsuka S,Hosoya J,Yamazaki M.PoLyInfo: polymer database for polymeric materials design. 2011 International Conference on Emerging Intelligent Data and Web Technologies; Tirana, Albania; 2011. pp. 22-9.

[119] Li M,Wang YG.Fast haar transforms for graph neural networks.Neural Netw2020;128:188-98

[120] Ma H,Rong Y.Cross-dependent graph neural networks for molecular property prediction.Bioinformatics2022;38:2003-9

[121] Mansimov E,Kang S.Molecular geometry prediction using a deep generative graph neural network.Sci Rep2019;9:20381 PMCID:PMC6938476

[122] Allotey J,Thiyagalingam J.Entropy-based active learning of graph neural network surrogate models for materials properties.J Chem Phys2021;155:174116

[123] Bertinetto C,Micheli A,Starita A.Evaluation of hierarchical structured representations for QSPR studies of small molecules and polymers by recursive neural networks.J Mol Graph Model2009;27:797-802

[124] Feinberg EN,Wu Z.PotentialNet for molecular property prediction.ACS Cent Sci2018;4:1520-30 PMCID:PMC6276035

[125] Withnall M,Engkvist O.Building attention and edge message passing neural networks for bioactivity and physical-chemical property prediction.J Cheminform2020;12:1 PMCID:PMC6951016

[126] Sun FY,Verma V. InfoGraph: Unsupervised and semi-supervised graph-level representation learning via mutual information maximization. International Conference on Learning Representations; Addis Ababa, Ethiopia; 2020. pp. 1-16. Available from: https://research.aalto.fi/en/publications/infograph-unsupervised-and-semi-supervised-graph-level-representa [Last accessed on 8 Jun 2023]

[127] Huang K,Glass LM,Sun J.SkipGNN: predicting molecular interactions with skip-graph networks.Sci Rep2020;10:21092 PMCID:PMC7713130

[128] Wang X,Jiang M,Zhang S.Molecule property prediction based on spatial graph embedding.J Chem Inf Model2019;59:3817-28

[129] Zhang Z,Zhou S.FraGAT: a fragment-oriented multi-scale graph attention model for molecular property prediction.Bioinformatics2021;37:2981-7 PMCID:PMC8479684

[130] Choi JY,Mehta K,Lupo Pasini M.Scalable training of graph convolutional neural networks for fast and accurate predictions of HOMO-LUMO gap in molecules.J Cheminform2022;14:70 PMCID:PMC9575242

[131] Choudhary K,Siderius DW,Mcdannald A.Graph neural network predictions of metal organic framework CO₂ adsorption properties.Comput Mater Sci2022;210:111388

[132] Frey NC,Jariwala D.Machine learning-enabled design of point defects in 2D materials for quantum and neuromorphic information processing.ACS Nano2020;14:13406-17

[133] Wang Z,Cai J,Li J.DeepTMC: A deep learning platform to targeted design doped transition metal compounds.Energy Stor Mater2022;45:1201-11

[134] Wang R,Zhang C,Yang M.Combining crystal graphs and domain knowledge in machine learning to predict metal-organic frameworks performance in methane adsorption.Micropor Mesopor Mat2022;331:111666

[135] Pablo-García S,Vargas-Hernández RA.Fast evaluation of the adsorption energy of organic molecules on metals via graph neural networks.Nat Comput Sci2023;3:433-42

[136] Tavazza F,Choudhary K.Uncertainty prediction for machine learning models of material properties.ACS Omega2021;6:32431-40 PMCID:PMC8655759

[137] Kwon Y,Choi YS.Uncertainty-aware prediction of chemical reaction yields with graph neural networks.J Cheminform2022;14:2 PMCID:PMC8750748

[138] Behler J.Generalized neural-network representation of high-dimensional potential-energy surfaces.Phys Rev Lett2007;98:146401

[139] Shapeev AV.Moment tensor potentials: a class of systematically improvable interatomic potentials.Multiscale Model Simul2016;14:1153-73

[140] Artrith N.An implementation of artificial neural-network potentials for atomistic materials simulations: Performance for TiO₂.Comput Mater Sci2016;114:135-50

[141] Wang H,Han J.DeePMD-kit: a deep learning package for many-body potential energy representation and molecular dynamics.Comput Phys Commun2018;228:178-84

[142] Unke OT,Sauceda HE.Machine learning force fields.Chem Rev2021;121:10142-86 PMCID:PMC8391964

[143] Dral PO.Quantum chemistry in the age of machine learning.J Phys Chem Lett2020;11:2336-47

[144] Yao N,Fu ZH.Applying classical, ab initio, and machine-learning molecular dynamics simulations to the liquid electrolyte for rechargeable batteries.Chem Rev2022;122:10970-1021

[145] Mai H,Chen D,Caruso RA.Machine learning for electrocatalyst and photocatalyst design and discovery.Chem Rev2022;122:13478-515

[146] Kovács DP,Kucera J.Linear atomic cluster expansion force fields for organic molecules: beyond RMSE.J Chem Theory Comput2021;17:7696-711 PMCID:PMC8675139

[147] Morrow JD,Deringer VL.How to validate machine-learned interatomic potentials.J Chem Phys2023;158:121501

[148] Behler J.Four generations of high-dimensional neural network potentials.Chem Rev2021;121:10037-72

[149] Deringer VL,Csányi G.A general-purpose machine-learning force field for bulk and nanostructured phosphorus.Nat Commun2020;11:5461 PMCID:PMC7596484

[150] Jinnouchi R,Karsai F,Bokdam M.Phase transitions of hybrid perovskites simulated by machine-learning force fields trained on the fly with bayesian inference.Phys Rev Lett2019;122:225701

[151] Margraf JT.Science-driven atomistic machine learning.Angew Chem Int Ed Engl2023;26:e202219170

[152] Freitas R.Uncovering the effects of interface-induced ordering of liquid on crystal growth using machine learning.Nat Commun2020;11:3260 PMCID:PMC7319977

[153] Schran C,Rowe P,Marsalek O.Machine learning potentials for complex aqueous systems made simple.Proc Natl Acad Sci U S A2021;118:e2110077118 PMCID:PMC8463804

[154] Anwar J.Uncovering molecular processes in crystal nucleation and growth by using molecular simulation.Angew Chem Int Ed Engl2011;50:1996-2013

[155] Caro MA,Koskinen J,Csányi G.Growth mechanism and origin of high sp^{3} content in tetrahedral amorphous carbon.Phys Rev Lett2018;120:166101

[156] Li J,An Q.Activating mobile dislocation in boron carbide at room temperature via al doping.Phys Rev Lett2023;130:116104

[157] Cao B,Sun A,Zhang T.Domain knowledge-guided interpretive machine learning: formula discovery for the oxidation behavior of ferritic-martensitic steels in supercritical water.J Mater Inf2022;2:4

[158] Li X,Dvornek N.BrainGNN: interpretable brain graph neural network for fMRI analysis.Med Image Anal2021;74:102233 PMCID:PMC9916535

[159] Ali A,Zakarya M.Exploiting dynamic spatio-temporal graph convolutional neural networks for citywide traffic flows prediction.Neural Netw2022;145:233-47

[160] Yang L,Zhang F.Multitask learning with graph neural network for travel time estimation.Comput Intell Neurosci2022;2022:6622734 PMCID:PMC8979717

[161] Wu X,Wang J,Guo Y.UBAR: user behavior-aware recommendation with knowledge graph.Knowl-Based Syst2022;254:109661

[162] Zhu J.A recommender for research collaborators using graph neural networks.Front Artif Intell2022;5:881704 PMCID:PMC9376356

[163] Gatta V, Moscato V, Postiglione M, Sperli G. An epidemiological neural network exploiting dynamic graph structured data applied to the COVID-19 outbreak.IEEE Trans Big Data2021;7:45-55

[164] Fritz C,Rügamer D.Combining graph neural networks and spatio-temporal disease models to improve the prediction of weekly COVID-19 cases in Germany.Sci Rep2022;12:3930 PMCID:PMC8913758