[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123587-en":3,"doc-seo-123587-105":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},123587,1099513958762,"Logic","https://ap-avatar.wpscdn.com/avatar/1000023916a998db790?x-image-process=image/resize,m_fixed,w_180,h_180&k=1784791008015729253",8,"Research & Report","Interpretable machine learning of amino acid patterns in proteins - a statistical ensemble approach","Explainable and interpretable unsupervised machine learning is used to reveal the structural information encoded in protein sequences. The study introduces an ensemble analysis that consolidates model interpretability and tests robustness by training restricted Boltzmann machines of equal size with different weight initializations. Optimal models compress five-amino-acid patterns at helix and sheet ends into a few bits and expose amino-acid-specific roles, including amphiphilicity markers and unexpected contributors.","arXiv :2303 . 15228v1 [ q-bio .BM] 27 Mar 2023  \nInterpretable machine learning of amino acid patterns in proteins: a statistical ensemble approach  \nAnna Braghetto, Enzo Orlandini, and Marco Baiesi  \nDipartimento di Fisica e Astronomia, Universit􀀒a di Padova, Via Marzolo 8, 35131, Padova, Italy and INFN, Sezione di Padova, Via Marzolo 8, 35131, Padova, Italy  \nExplainable and interpretable unsupervised machine learning helps understand the underlying structure of data. We introduce an ensemble analysis of machine learning models to consolidate their interpretation. Its application shows that restricted Boltzmann machines compress consistently into a few bits the information stored in a sequence of 􀀌ve amino acids at the start or end of 􀀋 -helices or 􀀌-sheets. The weights learned by the machines reveal unexpected properties of the amino acids and the secondary structure of proteins: (i) His and Thr have a negligible contribution to the amphiphilic pattern of 􀀋-helices; (ii) there is a class of 􀀋-helices particularly rich in Ala at their end; (iii) Pro occupies most often slots otherwise occupied by polar or charged amino acids, and its presence at the start of helices is relevant; (iv) Glu and especially Asp on one side, and Val, Leu, Iso, and Phe on the other, display the strongest tendency to mark amphiphilic patterns, i.e., extreme values of an e􀀋ective hydrophobicity, though they are not the most powerful (non) hydrophobic amino acids.  \nI. INTRODUCTION  \nVarious machine learning (ML) methods are applied to proteins [1{24] . For example, outstanding advancements have shown how ML can boost the prediction of protein native states [12{14] and complexes [15, 16] based only on amino acid sequences. However, the aim of several approaches is not to achieve a reliable (black box) tool for protein structure prediction but to get informative knowledge from the big data available for protein sequences and structures.  \nInterpretable ML [25, 26] focuses on understanding the cause of a model's decision and enhancing human capability to consistently predict the model's result. Interpretable ML versions are more complex and informative than standard statistical analysis and can improve our understanding of proteins [1{9] . In particular, they can detect patterns not emerging naturally from studying abundance and correlations of amino acids in secondary structures. Among the well-known patterns, for instance, there is the amphiphilic structure of several 􀀋-helices and 􀀌-sheets [27, 28], which are mostly (charged or) polar (P) on one side and nonpolar (N) on the other side. In an 􀀋 -helix, with pitch 􀀙 3:6 residues, the typical (non)polarity switch occurs every two residues. On the other hand, ina 􀀌-sheet, the three-dimensional alternation of the side chains takes place at every step. Hence an amphiphilic sequence would be, for example, PNPNP.  \nIn this work, we use a simple form of interpretable unsupervised ML, restricted Boltzmann machines (RBMs) [29{37], which allow extracting deep, nontrivial insight without losing the most transparent information on data statistics encoded in local biases. Conveniently, the weights and biases learned by RBMs can be visualized and easily interpreted. This established approach already revealed correlated amino acids within protein families [1], drug-target interactions [2], and correlations within DNA sequences [38, 39] .  \nA novelty of our work is a statistical ensemble approach to unsupervised ML, which improves the robustness of the 􀀌ndings. By training RBMs of the same size but with di􀀋erent weight initializations, we check whether they all converge to the same 􀀌nal set of learned weights. The maximally complex RBMs preserving this ensemble coherence are optimal, as they perform encoding of the correlations within data samples while providing stable and transparent information on the data.  \nWe show that our optimal RBMs perform extreme information compression to two or three bits, encoding the es","cbCaicqm6Owfbnvg","https://ap.wps.com/l/cbCaicqm6Owfbnvg","pdf",4080425,1,15,"English","en",105,"# Introduction\n## Interpretable and explainable unsupervised ML\n## Amphiphilic patterns in protein secondary structures\n## Restricted Boltzmann machines and interpretability\n## Ensemble approach and robustness rationale\n## Main findings on amino-acid contributions","[{\"question\":\"What is the goal of the ensemble analysis in this work?\",\"answer\":\"The approach consolidates interpretability across multiple unsupervised models by training restricted Boltzmann machines of the same size with different weight initializations, checking whether they converge to a consistent set of learned weights.\"},{\"question\":\"What do the optimal restricted Boltzmann machines achieve in terms of information compression?\",\"answer\":\"They perform extreme compression, encoding essential correlations between amino acids at the beginning or end of alpha-helices and beta-sheets into only two or three bits.\"},{\"question\":\"Which amino acids show notable roles in amphiphilic patterns according to the learned weights?\",\"answer\":\"The results indicate negligible contributions of His and Thr to the amphiphilic pattern of alpha-helices, Ala-enriched helices at their ends, and strong amphiphilic tendencies from Glu/Asp and from Val/Leu/Ile/Phe, even though other amino acids may show higher experimental hydrophobicity.\"}]","Interpretable machine learning of amino acid patterns in proteins - a statistical ensemble approach | PDF",1785817505,38,{"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},"interpretable-machine-learning-of-amino-acid-patterns-in-proteins-a-statistical-ensemble-approach","",{"@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/interpretable-machine-learning-of-amino-acid-patterns-in-proteins-a-statistical-ensemble-approach/123587/",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","2026-08-04",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},"What is the goal of the ensemble analysis in this work?","Question",{"text":75,"@type":76},"The approach consolidates interpretability across multiple unsupervised models by training restricted Boltzmann machines of the same size with different weight initializations, checking whether they converge to a consistent set of learned weights.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What do the optimal restricted Boltzmann machines achieve in terms of information compression?",{"text":80,"@type":76},"They perform extreme compression, encoding essential correlations between amino acids at the beginning or end of alpha-helices and beta-sheets into only two or three bits.",{"name":82,"@type":73,"acceptedAnswer":83},"Which amino acids show notable roles in amphiphilic patterns according to the learned weights?",{"text":84,"@type":76},"The results indicate negligible contributions of His and Thr to the amphiphilic pattern of alpha-helices, Ala-enriched helices at their ends, and strong amphiphilic tendencies from Glu/Asp and from Val/Leu/Ile/Phe, even though other amino acids may show higher experimental hydrophobicity.","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"]