[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119100-en":3,"doc-seo-119100-105":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},119100,5909877438554,"Maeve","https://ap-avatar.wpscdn.com/avatar/5600025385ad2bf12a7?_k=1778553567797529272",8,"Research & Report","Integration of generative machine learning with the heuristic crystal structure prediction code FUSE - Abstract","The prediction of new compounds via crystal structure prediction can reshape how the materials chemistry community discovers crystalline phases. Three broad routes exist: heuristic algorithms using established CSP codes, generative machine learning models that produce crystal structures directly, and mathematical optimisation that solves crystal structures exactly under constraints. This work combines heuristic and generative ML by using a generative model to create the starting population for a heuristic algorithm, and evaluates the approach on eight known and three hypothetical compounds. Results show faster per-structure computation and lower energies, and deliver an eleven-compound benchmark spanning diverse chemistry and complexity.","Integration of generative machine learning with the heuristic crystal structure prediction code FUSE  \nChristopher M. Collins 1,2 , Hasan M. Sayeed3 , George R. Darling 1 , John B. Claridge 1 ,  \nTaylor D. Sparks3 , Matthew J. Rosseinsky 1,2  \n1 Department of Chemistry, University of Liverpool, Crown Street, Liverpool, L69 7ZD, United Kingdom.  \n2 Leverhulme Research Centre for Functional Materials Design, Materials Innovation Factory, University of Liverpool, Crown Street, Liverpool, L69 7ZD, United Kingdom.  \n3 Department of Materials Science and Engineering, University of Utah, 122 Central Campus Dr, Salt Lake City, 84112, Utah, United States.  \nAbstract  \nThe prediction of new compounds via crystal structure prediction may transform how the materials chemistry community discovers new compounds. In the prediction of inorganic crystal structures there are three distinct classes of prediction; Performing crystal structure prediction via heuristic algorithms, using a range of established crystal structure prediction codes, an emerging community using generative machine learning models to predict crystal structures directly and the use of mathematical optimisation to solve crystal structures exactly. In this work, we demonstrate the combination of heuristic and generative machine learning, the use of a generative machine learning model to produce the starting population of crystal structures fora heuristic algorithm and discuss the benefits, demonstrating the method on eight known compounds with reported crystal structures and three hypothetical compounds. We show that the integration of machine learning structure generation with heuristic structure prediction results in both faster compute times per structure and lower energies. This work provides to the community a set of eleven compounds with varying chemistry and complexity that can be used as a benchmark for new crystal structure prediction methods as they emerge.  \nKeywords: Inorganic chemistry, Crystal structure prediction, Machine learning, Materials chemistry  \n1 Introduction  \nIn inorganic materials chemistry, the discovery of new materials can be guided by crystal structure prediction (CSP) . The use of CSP in the discovery of new compounds allows for experimental researchers to focus their efforts only on those compositions which are likely to yield new crystalline phases, greatly accelerating the speed with which they can be found[1, 2] . There are three main approaches to the crystal structure prediction problem. Firstly, well established codes based on heuristic algorithms which evolve crystal structures from a starting point, such as FUSE[3], which is based on a basin hopping algorithms, or USPEX[4] based upon genetic algorithms. Also within this class of structure search is the particle swarm optimisation method CALYPSO[5] . Related to this are random structure search methods which use randomly generated structures, constrained with a set of chemical rules, such as the code AIRSS[6] .  \nA second, recently emerging approach to the prediction of crystal is through the use of machine learning (ML) models. ML structure generation making use of now well-curated structure databases such as the Materials Project[7] or ICSD[8], to train models to rapidly generate plausible crystal structures for inorganic solids. These models include the use of graph neural networks[9], diffusion models[10] or large language models[11] . ML models have reached the point where they can efficiently generate large numbers of plausible crystal structures  \nfor a target composition. However, when such models are used far from their training data, or for non-trivial compositions they frequently fail to produce the correct structure, or a structure which can be relaxed into the ground state with a conventional chemistry calculation, for example with density functional theory (DFT) . A final, newly emerging approach is that of mathematical optimisation, where the structure prediction problem i","cbCaicl4WIWbw1Y0","https://ap.wps.com/l/cbCaicl4WIWbw1Y0","pdf",30604897,1,16,"English","en",105,"# Abstract\n# 1 Introduction\n## Crystal structure prediction approaches\n## Heuristic methods and local optimisation\n## Hybrid ML + heuristic strategy","[{\"question\":\"What three approaches to inorganic crystal structure prediction are discussed?\",\"answer\":\"The document describes heuristic-algorithm CSP codes, generative machine learning models for direct structure generation, and mathematical optimisation formulated to solve for the global minimum under constraints.\"},{\"question\":\"How does the proposed hybrid method combine generative ML with FUSE?\",\"answer\":\"It integrates a generative machine learning model to generate the starting population of crystal structures, replacing conventional random structure generation inside the heuristic CSP code FUSE.\"},{\"question\":\"What benefits does the integration of ML-generated structures provide?\",\"answer\":\"The method produces faster computation times per structure and lower energies compared with using conventional heuristic starting structures.\"}]","Integration of generative machine learning with the heuristic crystal structure prediction code FUSE - Abstract | PDF",1785722388,40,{"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},"integration-of-generative-machine-learning-with-the-heuristic-crystal-structure-prediction-code-fuse-abstract","",{"@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/integration-of-generative-machine-learning-with-the-heuristic-crystal-structure-prediction-code-fuse-abstract/119100/",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","2026-08-04","2026-08-03",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},"What three approaches to inorganic crystal structure prediction are discussed?","Question",{"text":76,"@type":77},"The document describes heuristic-algorithm CSP codes, generative machine learning models for direct structure generation, and mathematical optimisation formulated to solve for the global minimum under constraints.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the proposed hybrid method combine generative ML with FUSE?",{"text":81,"@type":77},"It integrates a generative machine learning model to generate the starting population of crystal structures, replacing conventional random structure generation inside the heuristic CSP code FUSE.",{"name":83,"@type":74,"acceptedAnswer":84},"What benefits does the integration of ML-generated structures provide?",{"text":85,"@type":77},"The method produces faster computation times per structure and lower energies compared with using conventional heuristic starting structures.","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},[94,98,102,106,111,116,120,123,128,131,135],{"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":29,"slug":119},7,"Healthcare","healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":107,"slug":138},19,"General","general"]