[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119766-en":3,"doc-seo-119766-105":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":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},119766,549758146520,"Patrick","https://ap-avatar.wpscdn.com/avatar/80002397d8c0411e94?_k=1775819394049821470",8,"Research & Report","Novel machine learning approaches revolutionize protein knowledge - Review","Breakthrough machine learning methods are reshaping structural biology by enabling accurate protein structure prediction and large-scale functional annotation. Techniques such as AlphaFold 2 (AF2) achieve model accuracy comparable to experimental structures, while protein language models and ultrafast structural aligners help validate and sustain annotation quality. The review explains how these advances close the growing gap between rapidly expanding protein sequences and the slower pace of structural/function characterization.","Trends in  \nBiochemical Sciences  \nOPEN ACCESS  \nReview  \nNovel machine learning approaches revolutionize protein knowledge  \nNicola Bordin  , 1 Christian Dallago  , 2,3 Michael Heinzinger  , 2,4 Stephanie Kim  , 5,6 Maria Littmann  2  \n,  \nClemens Rauer  , 1 Martin Steinegger  , 5,6,@ Burkhard Rost  , 2,7,8,@ and Christine Orengo  1,*,@  \nBreakthrough methods in machine learning (ML), protein structure prediction, and novel ultrafast structural aligners are revolutionizing structural biology. Obtaining accurate models of proteins and annotating their functions on a large scale is no longer limited by time and resources. The most recent method to be top ranked by the Critical Assessment of Structure Prediction (CASP) assessment, AlphaFold 2 (AF2), is capable of building structural models with an accuracy comparable to that of experimental structures. Annotations of 3D models are keeping pace with the deposition of the structures due to advancements in protein language models (pLMs) and structural aligners that help validate these transferred annotations. In this review we describe how recent developments in ML for protein science are making large-scale structural bioinformatics available to the general scientific community.  \nFrom protein sequence and structure to function through ML  \nThe number of experimentally determined, high-resolution structures deposited in the Protein Data Banki (PDB) [1] has grown immensely since its beginning in 1976, enabling research into biological mechanisms, and in turn the development of novel therapeutics and industrial applications. This growth is, however, outpaced exponentially by that of known protein sequences increasingly impacted by high-throughput metagenomic experiments which yield billions of entries per experiment. Closing the ever-increasing gap between protein sequence and annotations of structure and function is thus a desideratum in molecular and medical biology research.  \nMost proteins comprise two or more structural domains [2], that is, constituents with compact structures assumed to fold largely independently. Structural domains are often associated with specific functional roles [3], although functional sites can be formed from multiple domains [3] . These structural domains – often dubbed ‘folds’ – recur in nature [4], and have been estimated to be limited to a number in the order of thousands [5] . Folds resemble more the Plato’s allegory of the cave: more the image or idea or concept than the real object (Plato Politeia [6]); this image helps to map relations between proteins.  \nVarious resources emerged to classify domain structures in evolutionary families and fold groups (e.g. , SCOPii [7], CATHiii [8], SCOPeiv [9], and ECODv [10]), and these have saturated at about 5000 structural families and about 1300 folds over the past decade, despite structural genomics initiatives targeting proteins likely to have new folds [11] . As increasingly powerful sequence profile methods [12–14] have identified structural families in completely sequenced organisms (complete proteomes), studies suggest that up to 70% of all domains resemble those already classified in SCOP or CATH [3 , 15–17] . Trivially, the distribution of family size follows some power law: most families/folds are small or species-specific, but a few hundred are very highly populated, tend to be universal across species, and have important functions [8] . In parallel,  \nHighlights  \nTwo artificial intelligence (AI)-based methods for protein structure prediction, AlphaFold 2 and RoseTTAFold, increase dramatically the quality of structural modeling from sequence, nearing experimental accuracy.  \nProtein language models encode the written language of proteins, allowing for more accurate annotations and predictions than homology-based methods.  \nMost model organisms, neglected disease pathogens, and proteins with curated annotations have models available with varying quality, aiding wet-laboratory experiments targeting singl","cbCaimhr5yq54WyH","https://ap.wps.com/l/cbCaimhr5yq54WyH","pdf",3128675,1,15,"English","en",105,"# From protein sequence and structure to function through ML\n## Protein databases and the expanding sequence–structure gap\n## Structural domains, folds, and classification resources\n## Using AI to enable large-scale protein structural bioinformatics\n# Highlights\n## Protein structure prediction breakthroughs\n## Protein language models for annotation\n## Model organism coverage and curated datasets\n## Ultrafast alignment across protein space\n## Domain superfamily assignment from predicted models","[{\"question\":\"How do recent machine learning methods improve protein structure prediction?\",\"answer\":\"AlphaFold 2 and related approaches can generate structural models with accuracy comparable to experimental structures, greatly improving modeling from sequence information.\"},{\"question\":\"What role do protein language models play in protein science?\",\"answer\":\"Protein language models capture the “language” of proteins, enabling more accurate annotations and predictions than homology-based approaches.\"},{\"question\":\"Why are ultrafast structural aligners important?\",\"answer\":\"Ultrafast alignment tools traverse protein sequence and structure space to detect remote evolutionary relationships that older, slower methods could miss.\"}]","Novel machine learning approaches revolutionize protein knowledge - Review | PDF",1785726192,38,{"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},"novel-machine-learning-approaches-revolutionize-protein-knowledge-review","",{"@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/novel-machine-learning-approaches-revolutionize-protein-knowledge-review/119766/",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","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"How do recent machine learning methods improve protein structure prediction?","Question",{"text":75,"@type":76},"AlphaFold 2 and related approaches can generate structural models with accuracy comparable to experimental structures, greatly improving modeling from sequence information.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What role do protein language models play in protein science?",{"text":80,"@type":76},"Protein language models capture the “language” of proteins, enabling more accurate annotations and predictions than homology-based approaches.",{"name":82,"@type":73,"acceptedAnswer":83},"Why are ultrafast structural aligners important?",{"text":84,"@type":76},"Ultrafast alignment tools traverse protein sequence and structure space to detect remote evolutionary relationships that older, slower methods could miss.","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},[93,97,101,105,110,115,120,123,128,131,135],{"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":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":106,"slug":138},19,"General","general"]