{"paper_id":"0f9cbaee-2e82-4221-8a48-ecc1c44a7e15","body_text":"Regulatory Pathogenicity Is Mechanistically Heterogeneous: A Taxonomy of Activity-, Architecture-, and Coverage-Driven Blind Spots | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Regulatory Pathogenicity Is Mechanistically Heterogeneous: A Taxonomy of Activity-, Architecture-, and Coverage-Driven Blind Spots Sergey V. Boyko This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9090074/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background. Current variant interpretation tools assign pathogenicity along a single axis — typically sequence conservation or predicted functional impact. This conflation obscures mechanistically distinct classes of regulatory effect that require different computational approaches and different experimental validations. Whether regulatory pathogenicity decomposes into separable mechanistic axes, and how large the resulting blind spots are, has not been systematically assessed. Results. We propose a five-class taxonomy of regulatory pathogenicity: (A) activity-driven, where variants alter enhancer or promoter function detectable by reporter assays; (B) architecture-driven, where variants disrupt 3D chromatin contact topology detectable by structural simulation; (C) mixed, combining both mechanisms; (D) coverage gap, where current tools lack scoring capability; and (E) tissue-mismatch artifact, where apparent signals reflect incorrect tissue context. We classify 21 cases encompassing 30,318 ClinVar variants across 9 clinically important genomic loci using ARCHCODE, a loop-extrusion-based structural pathogenicity engine, integrated with VEP, CADD, MPRA cross-validation, and CRISPRi benchmarking. We show that 25 high-confidence and 29 candidate architecture-driven variants (Class B) are systematically missed by sequence-based tools: cross-locus weighted NMI(ARCHCODE, VEP) = 0.026; NMI at tissue-matched HBB = 0.495 (95% CI: 0.433–0.560). These variants cluster within 434 bp of tissue-matched enhancers (p = 2.51 × 10⁻³¹), 58-fold closer than activity-driven variants (25,138 bp), and return null results in both MPRA and CRISPRi screens — consistent with a contact-disruption rather than element-activity mechanism. An additional 207 coverage-gap variants (Class D) are unscored by VEP but detectable by structural simulation. Together, architecture-driven and coverage-gap variants account for 261 structural blind spots, of which 79.3% reflect tool absence (Class D) and 20.7% reflect true mechanistic orthogonality (Class B). Tissue-mismatch analysis (EXP-003) demonstrates that architecture-driven signal collapses by 700-fold in mismatched tissue (matched delta = 0.00357 vs. mismatch delta = 5.04 × 10⁻⁶), establishing tissue context as a necessary condition for Class B detection. A seven-locus tissue-match panel using ENCODE ChIP-seq data reveals four distinct outcome modes: positive amplification (SCN5A 1.37×, LDLR 1.43×), tail amplification (MLH1 2.0×), null (BRCA1 0.99×), and reverse effect (CFTR 0.60×, TERT 0.39×, TP53 0.18×), with reverse cases decomposing into overparameterization, enhancer loss, and enhancer dilution sub-mechanisms. Eight canonical cases from the literature — including TAD boundary disruption (Lupiáñez et al. 2015), insulated neighborhood disruption (Hnisz et al. 2016), and enhancer hijacking (Gröschel et al. 2014) — independently validate the taxonomy across limb malformations, leukemia, and medulloblastoma. Conclusions. Single-axis scoring is an inadequate abstraction for regulatory variant interpretation. Mechanistic decomposition reveals that architecture-driven pathogenicity — representing 20.7% of structural blind spots — requires dedicated 3D chromatin modeling that no current sequence-based tool provides. We propose that variant interpretation frameworks should explicitly assign mechanistic class before scoring, enabling targeted experimental validation and reducing systematic blind spots in clinical genetics. Epigenetics & Genomics regulatory variant interpretation chromatin architecture loop extrusion variant pathogenicity taxonomy non-coding variants 3D genome enhancer-promoter contacts tissue specificity blind spot analysis mechanistic decomposition Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-9090074\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":604184590,\"identity\":\"c09de1f7-b8bc-4725-b5fb-6b239181b6b1\",\"order_by\":0,\"name\":\"Sergey V. Boyko\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+0lEQVRIiWNgGAWjYNACgwQGBgnmAwc+ANls7MRrYUt8OAOkhZk4a0BaeIyNeUBsQlrk23sfv/hQkJbYP7stTdrm1zZ5PmYGxg8fc/A46cxxM8sZBjmJM+4cPiad23fbsI2ZgVly5jY8WiTS2Ix5DCoSN0ikpUnn9txmBGphY+bFo0V+/jM24z9gLTlm0pY9t+0JamG4wcb8mAHoMKAWY2OGH7cTCWoxOJPGxthjkGY840Za4sPehtvJbcyMzXj9It9+jPnDjz/Jsv0zkg8c+PHntu389uaDHz7icxgw7iTgTMY2MNmAVz0QMH9AsP8QUjwKRsEoGAUjEQAAFpJSNgkb7rwAAAAASUVORK5CYII=\",\"orcid\":\"\",\"institution\":\"Independent Researcher, Almaty, Kazakhstan\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Sergey\",\"middleName\":\"V.\",\"lastName\":\"Boyko\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2026-03-11 05:14:45\",\"currentVersionCode\":1,\"declarations\":{\"humanSubjects\":false,\"vertebrateSubjects\":false,\"conflictsOfInterestStatement\":false,\"humanSubjectEthicalGuidelines\":false,\"humanSubjectConsent\":false,\"humanSubjectClinicalTrial\":false,\"humanSubjectCaseReport\":false,\"vertebrateSubjectEthicalGuidelines\":false},\"doi\":\"10.21203/rs.3.rs-9090074/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-9090074/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":104780758,\"identity\":\"7005f813-eb3f-41f4-be58-24c2f9aecf95\",\"added_by\":\"auto\",\"created_at\":\"2026-03-17 07:53:49\",\"extension\":\"pdf\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":2912431,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"ARCHCODEtaxonomypaperRS.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-9090074/v1_covered_e5288017-8efd-412b-885d-eda9a17e006d.pdf\"}],\"financialInterests\":\"The authors declare no competing interests.\",\"formattedTitle\":\"\\u003cp\\u003eRegulatory Pathogenicity Is Mechanistically Heterogeneous: A Taxonomy of Activity-, Architecture-, and Coverage-Driven Blind Spots\\u003c/p\\u003e\",\"fulltext\":[],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":false,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":true,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":true,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":true,\"isPdf\":true,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"regulatory variant interpretation, chromatin architecture, loop extrusion, variant pathogenicity taxonomy, non-coding variants, 3D genome, enhancer-promoter contacts, tissue specificity, blind spot analysis, mechanistic decomposition\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-9090074/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-9090074/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eBackground. Current variant interpretation tools assign pathogenicity along a single axis — typically sequence conservation or predicted functional impact. This conflation obscures mechanistically distinct classes of regulatory effect that require different computational approaches and different experimental validations. Whether regulatory pathogenicity decomposes into separable mechanistic axes, and how large the resulting blind spots are, has not been systematically assessed.\\u003c/p\\u003e\\n\\u003cp\\u003eResults. We propose a five-class taxonomy of regulatory pathogenicity: (A) activity-driven, where variants alter enhancer or promoter function detectable by reporter assays; (B) architecture-driven, where variants disrupt 3D chromatin contact topology detectable by structural simulation; (C) mixed, combining both mechanisms; (D) coverage gap, where current tools lack scoring capability; and (E) tissue-mismatch artifact, where apparent signals reflect incorrect tissue context. We classify 21 cases encompassing 30,318 ClinVar variants across 9 clinically important genomic loci using ARCHCODE, a loop-extrusion-based structural pathogenicity engine, integrated with VEP, CADD, MPRA cross-validation, and CRISPRi benchmarking. We show that 25 high-confidence and 29 candidate architecture-driven variants (Class B) are systematically missed by sequence-based tools: cross-locus weighted NMI(ARCHCODE, VEP) = 0.026; NMI at tissue-matched HBB = 0.495 (95% CI: 0.433–0.560). These variants cluster within 434 bp of tissue-matched enhancers (p = 2.51 × 10⁻³¹), 58-fold closer than activity-driven variants (25,138 bp), and return null results in both MPRA and CRISPRi screens — consistent with a contact-disruption rather than element-activity mechanism. An additional 207 coverage-gap variants (Class D) are unscored by VEP but detectable by structural simulation. Together, architecture-driven and coverage-gap variants account for 261 structural blind spots, of which 79.3% reflect tool absence (Class D) and 20.7% reflect true mechanistic orthogonality (Class B). Tissue-mismatch analysis (EXP-003) demonstrates that architecture-driven signal collapses by 700-fold in mismatched tissue (matched delta = 0.00357 vs. mismatch delta = 5.04 × 10⁻⁶), establishing tissue context as a necessary condition for Class B detection. A seven-locus tissue-match panel using ENCODE ChIP-seq data reveals four distinct outcome modes: positive amplification (SCN5A 1.37×, LDLR 1.43×), tail amplification (MLH1 2.0×), null (BRCA1 0.99×), and reverse effect (CFTR 0.60×, TERT 0.39×, TP53 0.18×), with reverse cases decomposing into overparameterization, enhancer loss, and enhancer dilution sub-mechanisms. Eight canonical cases from the literature — including TAD boundary disruption (Lupiáñez et al. 2015), insulated neighborhood disruption (Hnisz et al. 2016), and enhancer hijacking (Gröschel et al. 2014) — independently validate the taxonomy across limb malformations, leukemia, and medulloblastoma.\\u003c/p\\u003e\\n\\u003cp\\u003eConclusions. Single-axis scoring is an inadequate abstraction for regulatory variant interpretation. Mechanistic decomposition reveals that architecture-driven pathogenicity — representing 20.7% of structural blind spots — requires dedicated 3D chromatin modeling that no current sequence-based tool provides. We propose that variant interpretation frameworks should explicitly assign mechanistic class before scoring, enabling targeted experimental validation and reducing systematic blind spots in clinical genetics.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Regulatory Pathogenicity Is Mechanistically Heterogeneous: A Taxonomy of Activity-, Architecture-, and Coverage-Driven Blind Spots\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2026-03-12 13:34:50\",\"doi\":\"10.21203/rs.3.rs-9090074/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"9dfb9fba-d183-4151-95de-2e4151bc6e61\",\"owner\":[],\"postedDate\":\"March 12th, 2026\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[{\"id\":64291977,\"name\":\"Epigenetics \\u0026 Genomics\"}],\"tags\":[],\"updatedAt\":\"2026-03-12T13:34:50+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2026-03-12 13:34:50\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-9090074\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-9090074\",\"identity\":\"rs-9090074\",\"version\":[\"v1\"]},\"buildId\":\"XKTyCvWXoU3ODBz1xrDgd\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}