{"id":11401,"date":"2025-03-17T08:00:00","date_gmt":"2025-03-17T07:00:00","guid":{"rendered":"https:\/\/haimagazine.com\/uncategorized\/a-revolution-in-predicting-protein-structures-how-artificial-intelligence-is-changing-the-game\/"},"modified":"2025-06-05T17:49:39","modified_gmt":"2025-06-05T15:49:39","slug":"a-revolution-in-predicting-protein-structures-how-artificial-intelligence-is-changing-the-game","status":"publish","type":"post","link":"https:\/\/haimagazine.com\/en\/uncategorized\/a-revolution-in-predicting-protein-structures-how-artificial-intelligence-is-changing-the-game\/","title":{"rendered":"\ud83d\udd12 A revolution in predicting protein structures: how artificial intelligence is changing the game"},"content":{"rendered":"<p>I&#8217;ve been fascinated for years by the mystery of protein folding\u2014the process where an amino acid chain takes on a three-dimensional shape, which determines its function. Traditional research methods, like X-ray crystallography or cryo-electron microscopy, require a lot of time and massive financial resources. The <strong>AlphaFold<\/strong> system created by Google DeepMind has revolutionized protein structure research.<\/p><h4 class=\"wp-block-heading\"><strong>Artificial intelligence is starting to play a significant role in natural sciences<\/strong><\/h4><p><strong><a href=\"https:\/\/alphafold.ebi.ac.uk\/\" target=\"_blank\" rel=\"noopener\"><mark style=\"background-color:#82D65E\" class=\"has-inline-color has-dark-gray-color\">AlphaFold<\/mark><\/a><\/strong> analyzes protein amino acid sequences and evolutional data. The model looks for related sequences in databases such as <mark style=\"background-color:#82D65E\" class=\"has-inline-color has-contrast-color\"><a href=\"https:\/\/www.uniprot.org\/\" target=\"_blank\" rel=\"noopener\">UniProt<\/a><\/mark>. This way, it&#8217;s possible to recognize homologous sequences and use this information to predict the protein&#8217;s spatial structure. The main element is <strong>multi-sequence analysis<\/strong> (MSA), which allows to recognize preserved structural patterns. Then, the <strong>Evoformer module<\/strong> processes these data, which allows to predict the interactions between amino acids and the structure of the protein&#8217;s spatial model. The <strong>AlphaFold 3<\/strong> version also features <strong>diffusion networks<\/strong>, which start from a random atom distribution, and then shape the stable structure considering the interactions with other biomolecules. The last stage is validating the model using indicators like <strong>pLDDT<\/strong>, which assesses the prediction accuracy within local structures, and <strong>pTM-score<\/strong>, which informs about the correctness of the protein&#8217;s global spatial organization, considering the relative position of its domain and structural elements.<\/p><figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"AlphaFold Server Demo - Google DeepMind\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/9ufplEgtq8w?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure><h4 class=\"wp-block-heading\"><strong>AI enhancing prediction accuracy<\/strong><\/h4><p>One variant of the system is <mark style=\"background-color:#ffffff\" class=\"has-inline-color\"> <\/mark><strong><a href=\"https:\/\/www.nature.com\/articles\/s41467-023-41721-9\" target=\"_blank\" rel=\"noopener\"><mark style=\"background-color:#82D65E\" class=\"has-inline-color has-dark-gray-color\">AlphaFold Multimer<\/mark><\/a><\/strong><mark style=\"background-color:#ffffff\" class=\"has-inline-color\">,<\/mark> designed to predict the structures of protein complexes. The model analyzes relationships between proteins and predicts interactions at the amino acid level. It features the <strong>ipTM-score<\/strong>, which assesses the quality of interactions between proteins more accurately. The future of AlphaFold Multimer is tied to the integration with AlphaFold 3, which could improve the accuracy of predictions. In the longer term, AlphaFold Multimer might become part of a broader AI ecosystem that could predict the behavior of complete biological systems and bring us closer to creating a virtual cell.<\/p><figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-4-3 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"AlphaFold Multimer | Folding complexes and multimers using AlphaFold2 | Full Tutorial\" width=\"500\" height=\"375\" src=\"https:\/\/www.youtube.com\/embed\/iGhdrPo6XQM?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure><p>Another tool is <strong><a href=\"https:\/\/www.nature.com\/articles\/s43588-024-00714-4\" target=\"_blank\" rel=\"noopener\"><mark style=\"background-color:#82D65E\" class=\"has-inline-color has-contrast-color\">MassiveFold<\/mark><\/a><\/strong>, a platform developed based on AlphaFold, designed for modeling complex protein structures, including multi-protein complexes and protein-ligand interactions. MassiveFold interprets these interactions correctly more often than AlphaFold does. The team developing MassiveFold also introduced an additional assessment indicator\u2014<strong>interface pLDDT<\/strong> (I-pLDDT), which is used to evaluate the quality of interaction areas between proteins.<\/p><p><strong><a href=\"https:\/\/www.science.org\/doi\/10.1126\/science.abj8754\" target=\"_blank\" rel=\"noopener\"><mark style=\"background-color:#82D65E\" class=\"has-inline-color has-contrast-color\">RoseTTAFold<\/mark><\/a><\/strong>, developed by the Institute for Protein Design at the University of Washington, is based on a <strong>deep neural network<\/strong> that simultaneously analyzes amino acid sequences, interactions between them, and spatial structure. Thanks to this, RoseTTAFold can reconstruct the shapes of proteins and enables the design of new structures, giving it an edge over other tools like AlphaFold.<\/p><figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-4-3 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"RoseTTAFold Full Tutorial | How to use RoseTTAFold2 for your research\" width=\"500\" height=\"375\" src=\"https:\/\/www.youtube.com\/embed\/cKfXUFoOQ0M?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure><h4 class=\"wp-block-heading\"><strong>Not just Google and DeepMind. Here are Meta AI&#8217;s protein language models<\/strong><\/h4><p><strong><a href=\"https:\/\/www.science.org\/doi\/10.1126\/science.ade2574\" target=\"_blank\" rel=\"noopener\"><mark style=\"background-color:#82D65E\" class=\"has-inline-color has-contrast-color\">ESMFold<\/mark><\/a><\/strong>, developed by <strong>Meta AI<\/strong>, utilizes <strong>protein language models<\/strong> (pLMs), which are trained on billions of sequences. This allows them to detect relationships between amino acids. The key benefit of ESMFold is its ability to instantly generate a protein structure, which significantly shortens analysis time. Future work on ESMFold will focus on <strong>integrating artificial intelligence methods with physicochemical protein models<\/strong> and on improving the modeling of protein complexes, currently being undertaken by AlphaFold Multimer.<\/p><figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"Meta ESM-2 Fold - AI faster than Alphafold 2\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/N-eisTvUYrk?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure><p><mark style=\"background-color:#82D65E\" class=\"has-inline-color has-contrast-color\"> <strong><a href=\"https:\/\/news.mit.edu\/2025\/ai-system-fragfold-predicts-protein-fragments-0220\" target=\"_blank\" rel=\"noopener\">FragFold<\/a><\/strong><\/mark>, a tool developed by scientists from the MIT Department of Biology, is also worth mentioning. By combining AlphaFold technology with new prediction algorithms, FragFold allows to predict how short protein fragments or peptides interact with larger target proteins.<\/p><p>I&#8217;m convinced that in the near future, the development of artificial intelligence will bring even more precise methods for predicting protein structures and their functions. AI-based tools not only solve complex problems in biology, but also open up new possibilities that seemed unreachable not so long ago.<\/p><p><\/p>","protected":false},"excerpt":{"rendered":"<p>The creators of AlphaFold revolutionized the expensive and time-consuming analysis of proteins, earning them the 2024 Nobel Prize in Chemistry. Yet this tool is today just one of many that utilize AI to model protein structures. What else can artificial intelligence do for biology and medicine?<\/p>\n","protected":false},"author":184,"featured_media":9228,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"rank_math_lock_modified_date":false,"footnotes":""},"categories":[],"tags":[707,748,715,749,752,753,747,750,751],"popular":[],"difficulty-level":[38],"ppma_author":[492],"class_list":["post-11401","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","tag-ai-5","tag-alphafold-2","tag-artificial-intelligence","tag-deepmind-2","tag-esmfold-2","tag-fragfold-2","tag-google-2","tag-massivefold-2","tag-rosettafold-2","difficulty-level-medium"],"acf":[],"authors":[{"term_id":492,"user_id":184,"is_guest":0,"slug":"anita-ciesielska","display_name":"dr hab. Anita Ciesielska","avatar_url":{"url":"https:\/\/haimagazine.com\/wp-content\/uploads\/2025\/04\/Yellow-and-Black-Simple-Professional-LinkedIn-Profile-Picture.png","url2x":"https:\/\/haimagazine.com\/wp-content\/uploads\/2025\/04\/Yellow-and-Black-Simple-Professional-LinkedIn-Profile-Picture.png"},"first_name":"","last_name":"","user_url":"","job_title":"","description":"Ekspertka AI w nauce i edukacji | Nauczycielka akademicka | Badaczka | Wydzia\u0142 Biologii i Ochrony \u015arodowiska | Uniwersytet \u0141\u00f3dzki"}],"_links":{"self":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/11401","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/users\/184"}],"replies":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/comments?post=11401"}],"version-history":[{"count":1,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/11401\/revisions"}],"predecessor-version":[{"id":11402,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/11401\/revisions\/11402"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/media\/9228"}],"wp:attachment":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/media?parent=11401"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/categories?post=11401"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/tags?post=11401"},{"taxonomy":"popular","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/popular?post=11401"},{"taxonomy":"difficulty-level","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/difficulty-level?post=11401"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/ppma_author?post=11401"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}