[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-128476-en":3,"doc-seo-128476-105":31,"detail-sidebar-cat-0-en-105":93},{"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":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},128476,13056712833777,"Logic","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Assessing accuracy of machine learning-based protein structure prediction model with molecular dynamics stabilization simulations - Master’s thesis","Advances in biotechnology enable synthetic proteins with tailored properties, yet the mapping between amino-acid sequence, 3D structure, and function remains incomplete. This thesis evaluates a machine learning protein structure predictor by running molecular dynamics stabilization simulations in Gromacs and testing predicted tandem-repeat proteins for stability. Because RMSD has interpretive limits, a new structural similarity measure, ρsc, is introduced. Simulation outcomes indicate many predicted proteins are unstable and can contain nonphysical overlapping-atom conformations, motivating hybrid computational approaches.","Assessing accuracy of machine learning-based protein structure prediction model with molecular dynamics stabilization simulations  \nCajsa Malm, 1901791  \nMaster’s thesis in physics Faculty of Science and Engineering at ˚Abo Akademi University  \nSupervisors:  \nProf. (Research) . Anssi Laukkanen Sr. Scientist. Antti Paajanen  \nTeam Leader (Research) . Antti Puisto Prof. Ronald ¨Osterbacka  \nMarch 29, 2023  \nAbstract  \nThe ability to synthetically create materials with desired properties has been greatly enhanced by advances in biotechnology. One important application of biotechnology is the design of synthetic proteins with the ability to fold in any configuration and predict their folding in advance. However, the connection between sequence, structure, and function is not yet fully understood, and there are endless combinations of amino acids that cannot be experimentally examined. Therefore, computational tools such as machine learning techniques are needed to guide material development.  \nThis thesis evaluates the accuracy of machine learning-based protein structure predictions by subjecting the predicted structures to molecular dynamics simulations using Gromacs software. The focus of the study is on synthetically constructed perfect tandem repeat proteins, and the goal is to test the stability of the predicted structures. The RMSD metric, often used to compare the structural similarity of proteins during molecular dynamics simulations, has limitations in its interpretation. To address this, a new measure of structural similarity called ρsc is proposed and used to assess the stability of the proteins.  \nThe results of the simulations show that many of the proteins generated by the machine learning model are unstable, with significant conformational changes observed. This suggests that the current model may not accurately predict the stability of all proteins. The predicted proteins also contained nonphysical structures with overlapping atoms. The study highlights the importance of combining machine learning approaches with other computational approaches to improve the accuracy of protein structure prediction.  \nIn conclusion, this study provides insights into the limitations of current machine learning models for protein structure prediction, and suggests the need for further research to better understand the underlying reasons for the observed instability. These findings could lead to improvements in protein design and prediction, ultimately leading to the creation of more advanced and functional materials with desired properties.  \nContents  \n1 Introduction 4  \n1.1 What is a protein and why is structure prediction important ..... 5  \n1.2 Machine learning in biology ....................... 6  \n1.3 Limitations in machine learning-based protein structure prediction .. 7  \n2 Theoretical basis for molecular dynamics simulations 9  \n2.1 Dihedral angles .............................. 10  \n2.2 Forces ................................... 12  \n2.3 Potential functions in MD simulations ................. 13  \n2.3.1 Potential energy functions for nonbonded interactions ..... 14  \n2.3.2 Potential energy functions for bonded interactions ....... 15  \n2.4 Initial energy minimization and equilibration .............. 17  \n2.5 MD parameters .............................. 20  \n3 Numerical Implementation and GROMACS Workflow 22  \n3.1 The Particle-Mesh Ewald method .................... 25  \n3.2 Integrators ................................. 27  \n3.2.1 The leapfrog integrator ...................... 28  \n3.2.2 Verlet integrator ......................... 28  \n3.2.3 Trotter decomposition ...................... 29  \n3.3 Temperature coupling .......................... 31  \n3.3.1 Berendsen temperature coupling ................. 32  \n3.3.2 Nos´e-Hoover temperature coupling ............... 32  \n3.3.3 Andersen thermostat ....................... 35  \n3.4 Pressure coupling ............................. 35  \n3.4.1 Berendsen pressure coupling .................","cbCaii2YQ43W3F6t","https://ap.wps.com/l/cbCaii2YQ43W3F6t","pdf",2505269,6,1,80,"English","en",105,"# Introduction\n## What is a protein and why is structure prediction important\n## Machine learning in biology\n## Limitations in machine learning-based protein structure prediction\n# Theoretical basis for molecular dynamics simulations\n## Dihedral angles\n## Forces\n## Potential functions in MD simulations\n### Potential energy functions for nonbonded interactions\n### Potential energy functions for bonded interactions\n## Initial energy minimization and equilibration\n## MD parameters\n# Numerical Implementation and GROMACS Workflow\n## The Particle-Mesh Ewald method\n## Integrators\n## Temperature coupling\n## Pressure coupling\n## Energy error\n## GROMACS Simulation Parameters: Choosing Force Field, Solvent Model, and Integration Settings\n# Results\n## Estimating the accuracy of machine learning model\n## Conclusion\n# Future outlooks","[{\"question\":\"What is the main objective of the thesis?\",\"answer\":\"To assess the accuracy of a machine learning-based protein structure prediction model by testing predicted synthetically constructed perfect tandem repeat proteins for stability using molecular dynamics simulations.\"},{\"question\":\"How are molecular dynamics simulations used in the evaluation?\",\"answer\":\"Predicted protein structures are subjected to molecular dynamics stabilization simulations in Gromacs, and their structural stability is analyzed during the simulation.\"},{\"question\":\"Why propose the ρsc measure instead of relying only on RMSD?\",\"answer\":\"RMSD is commonly used to compare structural similarity in molecular dynamics, but it has limitations in interpretation, motivating the introduction of a new similarity measure called ρsc.\"}]","Assessing accuracy of machine learning-based protein structure prediction model with molecular dynamics stabilization simulations - Master’s thesis | PDF",1786001268,202,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":88,"head_meta":90,"extra_data":92,"updated_unix":29},"assessing-accuracy-of-machine-learning-based-protein-structure-prediction-model-with-molecular-dynamics-stabilization-simulations-masters-thesis","",{"@graph":37,"@context":87},[38,55,70],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,49,52],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":48},"https://docshare.wps.com/document/","Document",2,{"item":50,"name":12,"@type":44,"position":51},"https://docshare.wps.com/document/research-report/",3,{"item":53,"name":13,"@type":44,"position":54},"https://docshare.wps.com/document/assessing-accuracy-of-machine-learning-based-protein-structure-prediction-model-with-molecular-dynamics-stabilization-simulations-masters-thesis/128476/",4,{"url":53,"name":13,"@type":56,"author":57,"headline":13,"publisher":59,"fileFormat":62,"inLanguage":24,"description":14,"dateModified":63,"datePublished":64,"encodingFormat":62,"isAccessibleForFree":65,"interactionStatistic":66},"DigitalDocument",{"name":9,"@type":58},"Person",{"url":42,"name":60,"@type":61},"DocShare","Organization","application/pdf","2026-08-25","2026-08-06",true,{"@type":67,"interactionType":68,"userInteractionCount":20},"InteractionCounter",{"@type":69},"ViewAction",{"@type":71,"mainEntity":72},"FAQPage",[73,79,83],{"name":74,"@type":75,"acceptedAnswer":76},"What is the main objective of the thesis?","Question",{"text":77,"@type":78},"To assess the accuracy of a machine learning-based protein structure prediction model by testing predicted synthetically constructed perfect tandem repeat proteins for stability using molecular dynamics simulations.","Answer",{"name":80,"@type":75,"acceptedAnswer":81},"How are molecular dynamics simulations used in the evaluation?",{"text":82,"@type":78},"Predicted protein structures are subjected to molecular dynamics stabilization simulations in Gromacs, and their structural stability is analyzed during the simulation.",{"name":84,"@type":75,"acceptedAnswer":85},"Why propose the ρsc measure instead of relying only on RMSD?",{"text":86,"@type":78},"RMSD is commonly used to compare structural similarity in molecular dynamics, but it has limitations in interpretation, motivating the introduction of a new similarity measure called ρsc.","https://schema.org",{"og:url":53,"og:type":89,"og:title":13,"og:site_name":60,"og:description":14},"article",{"robots":91,"canonical":53},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":94},[95,99,102,106,111,115,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":96,"show_sort_weight":97,"slug":98},"Story & Novel",90,"story-novel",{"id":48,"doc_module":4,"doc_module_name":47,"category_name":100,"show_sort_weight":22,"slug":101},"Literature","literature",{"id":54,"doc_module":4,"doc_module_name":47,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":47,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":112,"show_sort_weight":113,"slug":114},"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":47,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":47,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":47,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":47,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":47,"category_name":137,"show_sort_weight":107,"slug":138},19,"General","general"]