[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-detail-439770-en":59,"doc-seo-439770-105":82},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":5,"data":60},{"doc_id":61,"user_id":62,"nickname":63,"user_avatar":64,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":66,"doc_content":67,"file_id":68,"file_url":69,"file_type":70,"file_size":71,"view_count":72,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":73,"language":74,"language_code":75,"site_id":76,"html_lang":75,"table_of_contents":77,"faqs":78,"seo_title":79,"seo_description":66,"update_tm":80,"read_time":81},439770,1374404737137,"Adam","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d","Unsupervised discovery of clinical disease signatures using probabilistic independence","Objective: This study uses probabilistic independence to separate patient-specific disease sources from their clinical signatures within Electronic Health Record (EHR) data. Materials and Methods: The approach models each disease source as an unobserved root in a causal graph of observed EHR variables, defining signatures as downstream effects. Using 630,000 cross-sectional training instances from 269,099 longitudinal records, the method infers 2000 sources and signatures. Evaluation focuses on inferring and explaining benign versus malignant pulmonary nodules across 13,252 records via external references and literature comparisons.","Author Manuscript Author Manuscript Author Manuscript Author Manuscript  \n\n| | HHS Public Access\u003Cbr>Author manuscript\u003Cbr>J Biomed Inform. Author manuscript; available in PMC 2026 January 05. |\n| --- | --- |\n\nPublished in final edited form as:  \nJ Biomed Inform. 2025 June ; 166: 104837. doi:10.1016/j.jbi.2025.104837 .  \nUnsupervised discovery of clinical disease signatures using probabilistic independence  \nThomas A. Laskoa,b,* , William W. Steada, John M. Stilla, Thomas Z. Lib, Michael Kammera, Marco Barbero-Motaa, Eric V. Strobla,c, Bennett A. Landmana,b, Fabien MaldonadoaaVanderbilt University Medical Center, 1211 Medical Center Dr, Nashville, TN 37232, USA bVanderbilt University, 2301 Vanderbilt Pl, Nashville, TN 37235, USA  \nc University of Pittsburgh, 4200 Fifth Ave, Pittsburgh, PA 15260, USA  \nAbstract  \nObjective: This study uses probabilistic independence to disentangle patient-specific sources of disease and their signatures in Electronic Health Record (EHR) data.  \nMaterials and Methods: We model a disease source as an unobserved root node in the causal graph of observed EHR variables (laboratory test results, medication exposures, billing codes, and demographics), and a signature as the set of downstream effects that a given source has on those observed variables. We used probabilistic independence to infer 2000 sources and their signatures from 9195 variables in 630,000 cross-sectional training instances sampled at random times from  \n269,099 longitudinal patient records. We evaluated the learned sources by using them to infer and explain the causes of benign vs. malignant pulmonary nodules in 13,252 records, comparing the inferred causes to an external reference list and other medical literature. We compared models trained by three different algorithms and used corresponding models trained directly from the observed variables as baselines.  \nResults: The model recovered 92% of malignant and 30% of benign causes in the reference standard. Of the top 20 inferred causes of malignancy, 14 were not listed in the reference standard, but had supporting evidence in the literature, as did 11 of the top 20 inferred causes of benign nodules. The model decomposed listed malignant causes by an average factor of 5.5 and benign causes by 4.1, with most stratifying by disease course or treatment regimen. Predictive accuracy of  \nThis is an open access article under the CC BY license ([http://creativecommons.org/licenses/by/4.0/](http://creativecommons.org/licenses/by/4.0/)).  \n*Corresponding author., [tom.lasko@vanderbilt.edu](tom.lasko@vanderbilt.edu) (T.A. Lasko) .  \nDeclaration of competing interest  \nThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.  \nCRediT authorship contribution statement  \nThomas A. Lasko: Writing – review & editing, Writing – original draft, Validation, Supervision, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. William W. Stead: Writing – review & editing, Methodology, Conceptualization. John M. Still: Writing – review & editing, Software, Methodology. Thomas Z. Li: Writing – review & editing, Software, Methodology. Michael Kammer: Writing – review & editing, Validation, Methodology. Marco Barbero-Mota: Writing – review & editing, Methodology. Eric V. Strobl: Writing – review & editing, Writing – original draft, Methodology, Conceptualization. Bennett A. Landman: Writing – review & editing, Methodology. Fabien Maldonado: Writing – review & editing, Writing – original draft, Validation, Methodology.  \nAuthor Manuscript Author Manuscript Author Manuscript Author Manuscript  \nLasko et al. Page 2  \ncausal predictive models trained on source expressions (Random Forest AUC 0.788) was similar to (p = 0.058) their associational baselines (0.738) .  \nDiscussion: Most of the unrecovered causes were due to the rarity of the condition or lack of sufficien","cbCaiuEkQ0pYAJmB","https://ap.wps.com/l/cbCaiuEkQ0pYAJmB","pdf",1443689,3,33,"English","en",105,"# Abstract\n## Objective, Materials and Methods, Results, Discussion, Conclusion\n# Keywords\n## Background and significance","[{\"question\":\"What is the main goal of the study?\",\"answer\":\"The study aims to disentangle patient-specific disease sources and their signatures in EHR data using probabilistic independence, with downstream causal effects used to characterize signatures.\"},{\"question\":\"How does the model represent disease sources and signatures?\",\"answer\":\"It treats a disease source as an unobserved root node in the causal graph of observed EHR variables, and defines a signature as the downstream set of effects the source has on those observed variables.\"},{\"question\":\"How were the inferred causes evaluated in the paper?\",\"answer\":\"The learned sources were used to infer and explain causes of benign versus malignant pulmonary nodules in 13,252 records, comparing inferred causes to an external reference list and to supporting evidence in the medical literature.\"}]","Unsupervised discovery of clinical disease signatures using probabilistic independence | PDF",1790690142,83,{"code":4,"msg":83,"data":84},"ok",{"site_id":76,"language":75,"slug":85,"title":65,"keywords":86,"description":66,"schema_data":87,"social_meta":140,"head_meta":142,"extra_data":144,"updated_unix":145},"unsupervised-discovery-of-clinical-disease-signatures-using-probabilistic-independence","",{"@graph":88,"@context":139},[89,102,122],{"@type":90,"itemListElement":91},"BreadcrumbList",[92,96,98,100],{"item":93,"name":94,"@type":95,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":97,"name":9,"@type":95,"position":14},"https://docshare.wps.com/document/",{"item":99,"name":40,"@type":95,"position":72},"https://docshare.wps.com/document/research-report/",{"item":101,"name":65,"@type":95,"position":19},"https://docshare.wps.com/document/unsupervised-discovery-of-clinical-disease-signatures-using-probabilistic-independence/439770/",{"url":101,"name":65,"@type":103,"image":104,"author":109,"headline":65,"publisher":111,"fileFormat":114,"inLanguage":75,"description":66,"dateModified":115,"datePublished":116,"encodingFormat":114,"isAccessibleForFree":117,"interactionStatistic":118},"DigitalDocument",{"url":105,"@type":106,"width":107,"height":108},"https://docshare.wps.com/thumbnails/unsupervised-discovery-of-clinical-disease-signatures-using-probabilistic-independence/439770.png","ImageObject",300,407,{"name":63,"@type":110},"Person",{"url":93,"name":112,"@type":113},"DocShare","Organization","application/pdf","2026-10-01","2026-09-29",true,{"@type":119,"interactionType":120,"userInteractionCount":72},"InteractionCounter",{"@type":121},"ViewAction",{"@type":123,"mainEntity":124},"FAQPage",[125,131,135],{"name":126,"@type":127,"acceptedAnswer":128},"What is the main goal of the study?","Question",{"text":129,"@type":130},"The study aims to disentangle patient-specific disease sources and their signatures in EHR data using probabilistic independence, with downstream causal effects used to characterize signatures.","Answer",{"name":132,"@type":127,"acceptedAnswer":133},"How does the model represent disease sources and signatures?",{"text":134,"@type":130},"It treats a disease source as an unobserved root node in the causal graph of observed EHR variables, and defines a signature as the downstream set of effects the source has on those observed variables.",{"name":136,"@type":127,"acceptedAnswer":137},"How were the inferred causes evaluated in the paper?",{"text":138,"@type":130},"The learned sources were used to infer and explain causes of benign versus malignant pulmonary nodules in 13,252 records, comparing inferred causes to an external reference list and to supporting evidence in the medical literature.","https://schema.org",{"og:url":101,"og:type":141,"og:title":65,"og:site_name":112,"og:description":66},"article",{"robots":143,"canonical":101},"index,follow",{"doc_id":61,"site_id":76},1790792199]