[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-366251-105":59,"doc-detail-366251-en":131},{"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":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","novel-cancer-subtyping-method-guided-by-tumor-normal-sample-in-latent-space-of-transcriptomic-variational-autoencoder","Novel cancer subtyping method guided by tumor-normal sample in latent space of transcriptomic variational autoencoder","","Tumorigenesis is described as a microevolutionary process where heterogeneous cancer cells adapt within complex microenvironments. Existing tumor-omics analyses often ignore normal tissues, even though normal states can contextualize how tumors deviate. To close this gap, VaDTN (Variational Autoencoder–Derived Tumor-to-Normal) integrates transcriptomic data from tumor and normal samples into one latent space, then computes a distance from a normal reference for subtype discovery and survival stratification across multiple cancers.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/novel-cancer-subtyping-method-guided-by-tumor-normal-sample-in-latent-space-of-transcriptomic-variational-autoencoder/366251/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/novel-cancer-subtyping-method-guided-by-tumor-normal-sample-in-latent-space-of-transcriptomic-variational-autoencoder/366251.png","ImageObject",300,407,{"name":92,"@type":93},"Cart","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-30","2026-09-23",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",13,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"What problem does VaDTN address in existing tumor omics analyses?","Question",{"text":113,"@type":114},"It targets the common limitation that analyses rely mainly on tumor samples and overlook informative signals from normal tissues that help contextualize tumor deviations.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"How does VaDTN represent tumor and normal samples?",{"text":118,"@type":114},"It learns a shared low-dimensional latent representation from both tumor and normal transcriptomes using a variational autoencoder, positioning tumor samples relative to a normal tissue center.",{"name":120,"@type":111,"acceptedAnswer":121},"How are cancer subtypes and clinical relevance derived in VaDTN?",{"text":122,"@type":114},"Tumor distance embeddings from normal references are incorporated into clustering to form subtypes, and the resulting groups can show significant survival stratification in several cancer types.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},366251,1790217480,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":44,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145,"preview_page_count":19,"preview_urls":146},18829141979164,"https://eur-avatar.wpscdn.com/davatar_6f874abed73319feea01a86fa6f0fab8","[www. nature.com/scientificreports](www. nature.com/scientificreports)  \nOPEN  \nNovel cancer subtyping method guided by tumor-normal sample in latent space of transcriptomic variational autoencoder  \nHongzhi Wang, Yu Zhang, Dandan Zhang & Min Luo􀀍  \nTumorigenesis is a microevolutionary process in which heterogeneous tumor cells adapt within a complex microenvironment. Current tumor omics analyses often focus exclusively on tumor samples, overlooking the valuable insights that normal tissues can provide. To address this gap, we introduce VaDTN (Variational Autoencoder–Derived Tumor-to-Normal), a pan-cancer framework that integrates transcriptomic data from both tumor and normal samples into a unified latent space. By measuring each tumor’s “distance” from a normal reference within this latent space, VaDTN reveals subtle molecular shifts linked to tumor evolution and heterogeneity. We applied VaDTN to six representative cancers (SKCM, BRCA, LIHC, LUSC, STAD, and PAAD), identifying distinct subtypes characterized by unique transcriptional profiles. Notably, four cancer types (SKCM, BRCA, LIHC, and STAD) displayed significant survival stratification based on these subtype groupings, underscoring the clinical relevance of the distance-based approach. This reference-centered perspective thus provides a refined lens for dissecting intra-tumor diversity and guiding potential precision oncology strategies.  \nKeywords Representaion learning, Cancer subtype, Survival analysis  \nTumors are heterogeneous diseases driven by diverse molecular abnormalities, resulting in considerable variability both across patients and within individual tumors1–3. With the advent of high-throughput sequencing technologies4,5, pan-cancer studies have become a powerful method for integrating large-scale multi-omics data across different malignancies, aiming to pinpoint shared molecular features and pathways6,7. These analyses have identified key oncogenic drivers common to multiple tumor types while highlighting the complexity of inter-and intra-tumor heterogeneity. Capturing this complexity effectively remains a central challenge in cancer genomics8,9.  \nFurthermore, parallel advancements in other machine learning paradigms10, such as Topic Modeling11–13, have also demonstrated significant utility in identifying latent variables and hidden structures within complex biological datasets. Originally developed for text analysis, Topic Modeling has been increasingly applied in diverse biological contexts, including single-cell genomics and microbiome analysis, to uncover underlying biological states or cellular subpopulations14, 15. These methods, similar to Variational Autoencoders, aim to reduce dimensionality and reveal intrinsic patterns, making them relevant for comparison when seeking hidden disease subtypes. Incorporating such approaches has been shown to improve the resolution of biological insights16–18, for instance, by leveraging finer-grained class information during model training19.  \nMany pan-cancer studies either use normal tissues solely as a baseline to compare tumor expression patterns or exclude them entirely once differential expression analysis is complete20–22. While these approaches have yielded valuable insights, they often overlook the potential role of normal tissues in refining our understanding of how tumors deviate from physiological states23. By considering normal samples only peripherally, researchers may miss opportunities to contextualize tumor-related transcriptomic changes within a broader healthy reference—especially since each cancer type originates within a tissue-specific cellular and microenvironmental context23–25. Incorporating finer-grained tissue subtype labels during model training, as demonstrated in other contexts, could further enhance the resolution of such analyses. Consequently, a more comprehensive strategy that jointly models tumor and normal samples in a shared latent space could uncover subtle yet clinically ","cbCaiipGVtxK6Yxo","https://ap.wps.com/l/cbCaiipGVtxK6Yxo","pdf",1979585,"English","# Background\n## Tumor heterogeneity and pan-cancer omics\n## Topic modeling and latent-variable learning\n## Limitations of normal tissue usage\n# VaDTN Method\n## Joint latent representation from tumor and normal\n## Distance embedding to a normal tissue center\n## Distance-based clustering for subtypes\n# Applications and Findings\n## Subtypes across multiple cancer types\n## Survival stratification and clinical relevance","[{\"question\":\"What problem does VaDTN address in existing tumor omics analyses?\",\"answer\":\"It targets the common limitation that analyses rely mainly on tumor samples and overlook informative signals from normal tissues that help contextualize tumor deviations.\"},{\"question\":\"How does VaDTN represent tumor and normal samples?\",\"answer\":\"It learns a shared low-dimensional latent representation from both tumor and normal transcriptomes using a variational autoencoder, positioning tumor samples relative to a normal tissue center.\"},{\"question\":\"How are cancer subtypes and clinical relevance derived in VaDTN?\",\"answer\":\"Tumor distance embeddings from normal references are incorporated into clustering to form subtypes, and the resulting groups can show significant survival stratification in several cancer types.\"}]","Novel cancer subtyping method guided by tumor-normal sample in latent space of transcriptomic variational autoencoder | PDF",1790164398,23,[147,151,155,159],{"page_no":8,"url":148,"width":149,"height":150},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/3e2f117b37e342bc58ac019e17bfa417.webp",983,1291,{"page_no":14,"url":152,"width":153,"height":154},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/ebebdb50a6d951862dfd1713466e7b4a.webp",982,1290,{"page_no":81,"url":156,"width":157,"height":158},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/b9cbcdaf0f5fb5c383d47ab2fa2f320b.webp",981,1289,{"page_no":19,"url":160,"width":161,"height":162},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/b1adf090443984d91c5804b3f12c1b48.webp",980,1287]