[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126967-en":3,"doc-seo-126967-105":30,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},126967,687207024478,"Liam","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","Tell Machine Learning Potentials What They are Needed for - Simulation-Oriented Training Exemplified for Glycine","Machine learning potentials (MLPs) are widely used as efficient surrogates for potential energy surfaces (PES) in chemical simulations. This study systematically links common test-set error metrics to simulation accuracy using a full-dimensional global PES for glycine. Results show that test errors do not unambiguously predict performance across tasks including relative conformer energies, barriers, vibrational levels, and zero-point vibrational energies. A simulation-oriented training strategy improves MLP quality and passes diffusion Monte Carlo validation.","Tell Machine Learning Potentials What They are Needed for: SimulationOriented Training Exemplified for  \nGlycine  \nFuchun Ge, 1† Ran Wang, 1† Chen Qu,2 Peikun Zheng, 1,3Apurba Nandi,4,5 Riccardo Conte,6 Paul L. Houston,7 Joel M. Bowman,4* Pavlo O. Dral 1*  \n1State 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  \n2Independent Researcher, Toronto, Ontario M9B0E3, Canada  \n3Present address: Department of Chemistry, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, United States  \n4Department of Chemistry and Cherry L. Emerson Center for Scientific Computation, Emory University, Atlanta, Georgia, 30322, United States  \n5Department of Physics and Materials Science, University of Luxembourg, Luxembourg City L-1511, Luxembourg  \n6Dipartimento di Chimica, Università degli Studi di Milano, via Golgi 19, 20133 Milano, Italy  \n7Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York 14853, United States; Department of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, United States  \n[E-mails:](E-mails: jmbowma@emory.edu)[ ](E-mails: jmbowma@emory.edu)[jmbowma@emory.edu](E-mails: jmbowma@emory.edu); [dral@xmu.edu.cn](dral@xmu.edu.cn)  \n†Equal contribution  \nAbstract  \nMachine learning potentials (MLPs) are widely applied as an efficient alternative way to represent potential energy surfaces (PES) in many chemical simulations. The MLPs are often evaluated with the root-mean-square errors on the test set drawn from the same distribution asthe 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 the diffusion Monte Carlo simulations.  \nTOC Graphic  \nPotential energy surface (PES) 1-5 is one of the most essential concepts in computational chemistry, which can be abstracted as a multivariable function of the positions of nuclei with the output indicating the potential energy profile6, 7. The negative gradients ofthis function are the forces that atoms experience. PES is a potent instrument that allows us to understand molecular structures8, their stability9-11, and reactivity governed by nuclear dynamics12-18. PES is extensively utilized for the exploration of the conformational landscape and reaction mechanisms19-34.  \nDespite PES being an essential prerequisite to perform many chemical simulations, the construction often becomes a bottleneck in the application: highly accurate ab initio electronic structure methods offer reliable potential energies but have high computationally cost and scaling, while low-level empirical or semiempirical methods are fast but less robust. Fortunately, this conflict can be tackled by applying machine learning (ML) techniques to return an analytic functional form that mimics the PES, resulting in what is called machine learning potentials (MLPs) .35-51 By serving as an efficient replacement for quantum mechanically generated potential energies (and gradients), MLPs have developed rapidly over the past decade and various algorithms emerged. MLPs based on linear regression,49, 52 using permutationally invariant polynomials,50, 51 neural networks (NN),44, 47, 53 and ","cbCaiiw0lTKGeBar","https://ap.wps.com/l/cbCaiiw0lTKGeBar","pdf",975506,1,26,"English","en",105,"# Abstract\n## Background: PES and the need for MLPs\n## Limits of test-set metrics (RMSE/MAE)\n## Simulation-oriented validation approach\n## Study aim and glycine PES application","[{\"question\":\"Why are machine learning potentials used in chemical simulations?\",\"answer\":\"MLPs provide an efficient analytic representation of potential energy surfaces, replacing expensive quantum-mechanical energy and gradient calculations in simulations.\"},{\"question\":\"What problem arises when evaluating MLPs only with test-set RMSE/MAE?\",\"answer\":\"Test-set metrics can be non-representative and even misleading because they are dependent on the test data distribution and provide only overall average error rather than task-specific reliability.\"},{\"question\":\"How does the study evaluate MLP quality in a simulation-oriented manner?\",\"answer\":\"It directly examines the correctness of simulations performed with MLPs across multiple task types, including relative conformer energies, barriers, vibrational levels, and zero-point vibrational energies, and further validates with diffusion Monte Carlo.\"}]","Tell Machine Learning Potentials What They are Needed for - Simulation-Oriented Training Exemplified for Glycine | PDF",1785935954,66,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"tell-machine-learning-potentials-what-they-are-needed-for-simulation-oriented-training-exemplified-for-glycine","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/tell-machine-learning-potentials-what-they-are-needed-for-simulation-oriented-training-exemplified-for-glycine/126967/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-22","2026-08-05",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"Why are machine learning potentials used in chemical simulations?","Question",{"text":76,"@type":77},"MLPs provide an efficient analytic representation of potential energy surfaces, replacing expensive quantum-mechanical energy and gradient calculations in simulations.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What problem arises when evaluating MLPs only with test-set RMSE/MAE?",{"text":81,"@type":77},"Test-set metrics can be non-representative and even misleading because they are dependent on the test data distribution and provide only overall average error rather than task-specific reliability.",{"name":83,"@type":74,"acceptedAnswer":84},"How does the study evaluate MLP quality in a simulation-oriented manner?",{"text":85,"@type":77},"It directly examines the correctness of simulations performed with MLPs across multiple task types, including relative conformer energies, barriers, vibrational levels, and zero-point vibrational energies, and further validates with diffusion Monte Carlo.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]