[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-127141-en":3,"doc-seo-127141-105":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},127141,687207022233,"Riley","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Unraveling dynamic protein structures by two-dimensional infrared spectra with a pretrained machine learning model","Dynamic protein structures underpin biological function, yet connecting two-dimensional infrared (2DIR) spectral signals to rapid conformational evolution remains challenging and typically requires extensive expert analysis. This work presents a pretrained machine learning model trained on ~204,300 2DIR spectra to learn a spectrum–structure correlation. The model accurately predicts dynamic secondary-structure content changes and transfers across real folding trajectories spanning microseconds to milliseconds, while providing attention-based spectral explanations for conformational dynamics in native environments.","UC Irvine  \nUC Irvine Previously Published Works  \nTitle  \nUnraveling dynamic protein structures by two-dimensional infrared spectra with apretrained machine learning model.  \nPermalink  \n[https://escholarship.org/uc/item/9618v6hv](https://escholarship.org/uc/item/9618v6hv)  \nJournal  \nProceedings of the National Academy of Sciences, 121(27)  \nAuthors  \nWu, Fan  \nHuang, Yan Yang, Guokunet al.  \nPublication Date  \n2024-07-02  \nDOI  \n10.1073/pnas.2409257121  \nPeer reviewed  \n[eScholarship.org](eScholarship.org) Powered by the California Digital Library  \nUniversity of California  \nUnraveling dynamic protein structures by two-dimensional infrared spectra with a pretrained machine learning model  \nFan Wua,1, Yan Huanga,1, Guokun Yanga,1, Sheng Yeb,2, Shaul Mukamelc,2, and Jun Jianga,2  \nContributed by Shaul Mukamel; received May 9, 2024; accepted May 28, 2024; reviewed by Gregory D. Scholes and Martin T. Zanni  \nDynamic protein structures are crucial for deciphering their diverse biological functions. Two-dimensional infrared (2DIR) spectroscopy stands as an ideal tool for tracing rapid conformational evolutions in proteins. However, linking spectral characteristics to dynamic structures poses a formidable challenge. Here, we present apretrained machine learning model based on 2DIR spectra analysis. This model has learned signal features from approximately 204,300 spectra to establish a “spectrum-structure” correlation, thereby tracing the dynamic conformations of proteins. It excels in accurately predicting the dynamic content changes of various secondary structures and demonstrates universal transferability on real folding trajectories spanning timescales from microseconds to milliseconds. Beyond exceptional predictive performance, the model offers attention-based spectral explanations of dynamic conformational changes. Our 2DIR-based pretrained model is anticipated to provide unique insights into the dynamic structural information of proteins in their native environments.  \nultrafast spectroscopy | protein dynamics | machine learning  \nProtein structures are pivotal for elucidating their diverse biological functions. Significant experimental advancements have been made in the determination of protein structure (1–3) . In recent years, AI has shown promising success in determining the lowest-energy state of proteins (4–18) . Tools like AlphaFold2 (4, 5) and RoseTTAFold (6) can predict the three-dimensional structures of proteins from their amino acid sequences, while the integration of message passing neural network (MPNN) supplements the predictive capability of protein assemblies (8). The latest generative models can sample a broad variety of protein structures based on desired properties (13–16) . These advancements have deepened our understanding of the lowest-energy static protein structures. Given that the dynamic characteristics of proteins ultimately shape their biological functions (19), integrating conformational dynamics information into machine learning (ML) training is therefore crucial for identifying dynamic protein structures that are relevant to biological processes (20–22) .  \nOptical signals offer a unique window into protein dynamic responses. Two-dimensional infrared (2DIR) spectroscopy, based on femtosecond pulse sequences, has proven to be a powerful tool for determining protein structure and provides snapshots of protein folding events (23–30). However, unraveling dynamic protein structures from a series of 2DIR spectra present a formidable task, which typically requires days or weeks of manual analysis by a trained expert. Recent efforts in applying ML methods to extract structural information from spectroscopic signals (31–35) have paved the way for the potential of tracing protein dynamics. Therefore, it is imperative to develop data-driven ML protocols for automatically establishing correlations between protein 2DIR spectra and their dynamic conformations.  \nHere, we introduce a ML pretrained model ut","cbCaihUHoKuqZS7E","https://ap.wps.com/l/cbCaihUHoKuqZS7E","pdf",5237728,1,9,"English","en",105,"# Abstract\n# Significance\n# Introduction\n# Optical signals and 2DIR spectroscopy\n# Methods and ML model\n## Pretraining and spectrum-structure correlation\n## Transferability across folding trajectories\n# Results\n## Overall schematic and ML model architecture\n# Author affiliations","[{\"question\":\"Why is 2DIR spectroscopy important for studying protein dynamics?\",\"answer\":\"2DIR spectroscopy provides a powerful window into rapid conformational changes by capturing protein folding events as spectral signatures. It enables tracing dynamic evolution that static structure methods cannot directly reveal.\"},{\"question\":\"How does the pretrained machine learning model connect spectra to protein structure?\",\"answer\":\"The model learns signal features from approximately 204,300 2DIR spectra and establishes a spectrum–structure correlation. This learned mapping enables prediction of dynamic secondary-structure content.\"},{\"question\":\"What evidence of generalization and interpretability is reported?\",\"answer\":\"The model shows universal transferability to real folding trajectories from microseconds to milliseconds. It also offers attention-based spectral explanations that support interpreting dynamic conformational changes from the original spectra.\"}]","Unraveling dynamic protein structures by two-dimensional infrared spectra with a pretrained machine learning model | PDF",1785937144,23,{"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},"unraveling-dynamic-protein-structures-by-two-dimensional-infrared-spectra-with-a-pretrained-machine-learning-model","",{"@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/unraveling-dynamic-protein-structures-by-two-dimensional-infrared-spectra-with-a-pretrained-machine-learning-model/127141/",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","2026-08-05",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},"Why is 2DIR spectroscopy important for studying protein dynamics?","Question",{"text":75,"@type":76},"2DIR spectroscopy provides a powerful window into rapid conformational changes by capturing protein folding events as spectral signatures. It enables tracing dynamic evolution that static structure methods cannot directly reveal.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the pretrained machine learning model connect spectra to protein structure?",{"text":80,"@type":76},"The model learns signal features from approximately 204,300 2DIR spectra and establishes a spectrum–structure correlation. This learned mapping enables prediction of dynamic secondary-structure content.",{"name":82,"@type":73,"acceptedAnswer":83},"What evidence of generalization and interpretability is reported?",{"text":84,"@type":76},"The model shows universal transferability to real folding trajectories from microseconds to milliseconds. It also offers attention-based spectral explanations that support interpreting dynamic conformational changes from the original spectra.","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},[93,97,101,105,110,115,120,123,127,130,134],{"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":121,"slug":122},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},"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"]