[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-124224-en":3,"doc-seo-124224-105":30,"detail-sidebar-cat-0-en-105":91},{"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":4,"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},124224,1099514067438,"River Wang","https://ap-avatar.wpscdn.com/avatar/100002539ee87300030?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780474512215547542",8,"Research & Report","Extending atomic decomposition and many-body representation with a chemistry-motivated approach to machine learning potentials","Machine learning potentials for condensed-phase simulations often trade chemical interpretability in energy decomposition against computational efficiency. This paper introduces a chemistry-motivated framework that combines monomer-centered representation with feed-forward neural networks. The total energy is expressed as a sum of chemically meaningful monomeric energies whose descriptors come from one-body and two-body permutationally invariant polynomial interactions. Systematic evaluations on gas-phase water trimer, liquid water, methane–water clusters, and liquid carbon dioxide demonstrate improved accuracy, efficiency, and flexibility for large-scale quantum and classical simulations.","nature computational science  \nArticle [https://doi.org/10.1038/s43588-025-00790-0](https://doi.org/10.1038/s43588-025-00790-0)  \nExtending atomic decomposition and many-body representation with a chemistry-motivated approach to machine learning potentials  \nReceived: 16 August 2024  \nAccepted: 13 March 2025  \n\n| Published online: xx xx xxxx\u003Cbr> |\n| --- |\n|  Check for updates |\n\nQi Yu  1,2 , Ruitao Ma1, Chen Qu3, Riccardo Conte  4, Apurba Nandi5,  \nPriyanka Pandey6, Paul L. Houston7, Dong H. Zhang  8 & Joel M. Bowman  6  \nMost widely used machine learning potentials for condensed-phase applications rely on many-body permutationally invariant polynomial or atom-centered neural networks. However, these approaches face challenges in achieving chemical interpretability in atomistic energy decomposition and fully matching the computational efficiency of traditional force fields. Here we present a method that combines aspects of both approaches and balances accuracy and force-field-level speed. This method utilizesa monomer-centered representation, where the potential energy is decomposed into the sum of chemically meaningful monomeric energies. The structural descriptors of monomers are described by one-body and two-body effective interactions, enforced by appropriate sets of permutationally invariant polynomialsas inputs to the feed-forward neural networks. Systematic assessments of models for gas-phase water trimer, liquid water, methane–water cluster and liquid carbon dioxide are performed. The improved accuracy, efficiency and flexibility of this method have promise for constructing accurate machine learning potentialsand enabling large-scale quantum and classical simulations for complex molecular systems.  \nComputational simulations of molecular systems are essential for understanding complex processes in chemistry, biology and material sciences. A key challenge in both quantum and classical simulations is the extensive computations required for potential energy and force evaluations given molecular configurations. Direct ab initio calculations using accurate electronic-structure methods such as the ‘gold standard’coupled cluster theory with single, double and perturbative triple excitations, CCSD(T)1, are ideal. However, it quickly becomes prohibitive for systems with more than 15 atoms. Although density  \nfunctional theory is widely used inabinitio molecular dynamics (MD) simulations due to its relative efficiency, its limited accuracy and still unfavorable computational scaling present challenges for long-time simulations of large and complex systems.  \nOver the past two decades, machine learning potentials (MLPs) have emerged asa promising approach to enable efficient and accurate computational simulations2–24. For high-dimensional systems with tens of thousands atoms, such as condensed-phase water, anatomistic representation of the potentialisa popular choice13:  \n1Department of Chemistry, Fudan University, Shanghai, China. 2Shanghai Innovation Institute, Shanghai, China. 3Independent Researcher, Toronto, Ontario, Canada. 4Dipartimento di Chimica, Università degli Studi di Milano, Milan, Italy. 5Department of Physics and Materials Science, University of Luxembourg, Luxembourg City, Luxembourg. 6Department of Chemistry and Cherry L. Emerson Center for Scientific Computation, Emory University, Atlanta, GA, USA. 7Department of Chemistry and Chemical Biology, Cornell University, Ithaca, NY, USA. 8State Key Laboratory of Molecular Reaction  \nDynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China. [e-mail: qi_yu@fudan.edu.cn](e-mail: qi_yu@fudan.edu.cn)  \nNature Computational Science  \nNatom  \nEtotal = ∑i Ei,atomic , (1)  \nwhere the total potential energy of the system Etotal is decomposed asthe sum of atomic local energies Ei,atomic over all Natom atoms, where i represents the index of each atom in the system. This representation has been widely applied in various MLPs. Typical examples incl","cbCailVegXfeqX4f","https://ap.wps.com/l/cbCailVegXfeqX4f","pdf",1634366,1,12,"English","en",105,"# Introduction\n## Computational simulation challenge\n## Limits of quantum and classical methods\n## Machine learning potentials overview\n# Method\n## Monomer-centered energy decomposition\n## Permutationally invariant polynomial descriptors\n## Neural network architecture inputs\n# Evaluation\n## Gas-phase water trimer\n## Liquid water\n## Methane–water clusters\n## Liquid carbon dioxide","[{\"question\":\"What key problem does the proposed method address in existing machine learning potentials?\",\"answer\":\"It aims to improve chemical interpretability of atomistic energy decomposition while maintaining force-field-level computational efficiency.\"},{\"question\":\"How is the potential energy represented in the proposed framework?\",\"answer\":\"The total potential energy is decomposed into a sum of monomer-centered, chemically meaningful monomeric energies.\"},{\"question\":\"Which systems are used to assess the models in the study?\",\"answer\":\"The models are evaluated on gas-phase water trimer, liquid water, methane–water clusters, and liquid carbon dioxide.\"}]","Extending atomic decomposition and many-body representation with a chemistry-motivated approach to machine learning potentials | PDF",1785821097,30,{"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":86,"head_meta":88,"extra_data":90,"updated_unix":28},"extending-atomic-decomposition-and-many-body-representation-with-a-chemistry-motivated-approach-to-machine-learning-potentials","",{"@graph":36,"@context":85},[37,54,68],{"@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/extending-atomic-decomposition-and-many-body-representation-with-a-chemistry-motivated-approach-to-machine-learning-potentials/124224/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-04",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What key problem does the proposed method address in existing machine learning potentials?","Question",{"text":75,"@type":76},"It aims to improve chemical interpretability of atomistic energy decomposition while maintaining force-field-level computational efficiency.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How is the potential energy represented in the proposed framework?",{"text":80,"@type":76},"The total potential energy is decomposed into a sum of monomer-centered, chemically meaningful monomeric energies.",{"name":82,"@type":73,"acceptedAnswer":83},"Which systems are used to assess the models in the study?",{"text":84,"@type":76},"The models are evaluated on gas-phase water trimer, liquid water, methane–water clusters, and liquid carbon dioxide.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":29,"slug":121},"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]