state

Wiki State

Persistent operational context for the extrafollicular B cell dynamics in dengue literature review. Read this at the start of every session. Update it after every session.

---

Current Focus

[2026-06-29] Morra2018 ingested — severity axis now has both pillars. Added Morra2018 - Defining Warning Signs and Severe Dengue (Rev Med Virol; PRISMA review of 44 WHO-2009 studies) as the second source on the severity axis. 22 sources / 107 pages. Where Narvaez2011 established between-scheme heterogeneity (WHO-1997 vs WHO-2009, κ=0.25), Morra establishes within-scheme heterogeneity: across 44 WHO-2009 studies, only 2 of 16 warning/severe signs (liver enlargement; liver involvement) reach a consensus definition — both WHO-2009-predefined — while “shock” alone is defined 23 different ways. The two stack: a severity-stratified claim must carry not just which scheme but, ideally, which operational definitions produced it. Done with sub-agents (1 propagation mapper + 2 parallel drafters), mirroring the Narvaez workflow. Tight propagation (curator’s call): concept hub Dengue Severity Classification (sources 1→2) + 1 Notable Findings entry + 2 curated source links (GarciaBates2013, GodoyLozano2016); corpus-wide backlink still deferred. Accuracy guard held: Morra’s “73.0% vs 93.4% specificity” is attributed to Macedo et al (borrowed, different cohort/gold-standard than Narvaez’s 78.5%) — explicitly not a contradiction; Morra pools no diagnostic accuracy of its own.

[2026-06-29] Narvaez2011 ingested — severity-axis anchor created. New concept hub Dengue Severity Classification (WHO-1997 DF/DHF/DSS vs WHO-2009 Dengue±Warning Signs/Severe Dengue) + source page; 21 sources / 106 pages. Cross-cutting lesson for the whole wiki: severity associations are scheme-dependent (Narvaez κ=0.25 between schemes; the DENV-2→DHF/DSS signal vanishes under the revised scheme) → every severity-stratified claim must carry the scheme that produced it. Done with sub-agents (1 propagation mapper + 2 page drafters). Curated propagation to 5 high-value pages only; full backlinking deferred to a lint (new Watch Item). The pending severity-scheme decision now has its in-wiki evidentiary anchor.

[2026-06-14] New SOP page — DONE. Created DN2 Panel - Staining, Compensation, and Gating Protocol — full bench-to-gate operational SOP (blood prep → controls → FlowJo comp → FMO gating), worked-example-driven, with an explicit limitations section. Scope-fenced against Compensation and FMO Controls (owns the numbers) and DN2 Gating Strategy (owns the gating hierarchy); cross-linked from both.

[2026-06-14] Direction — pivot COMPLETE. High-level direction: plasmablasts + atypical/age-associated B cells and their association with autoantibodies and neutralizing antibodies in dengue. The wiki’s spine is reframed from “extrafollicular response” to atypical/age-associated B cells + plasmablasts, with the EF pathway as a generating mechanism under them. A sibling bridge-wiki/ (synthesis layer) connects this wiki’s cells to the dengue-wiki/ antibody/ANA content — see ../bridge-wiki/state.md and memory [[three-wiki-architecture]]. The reframe is fully executed across pages, index, state, and CLAUDE.md (H1/mission/Domain Context now anchor on the hybrid term “Atypical (DN)” — curator’s choice; the formerly-gated identity wording is resolved). [2026-06-14] Lamprinou2026 ingested (ABC↔DN identity opinion piece; new [[Age-Associated B Cell]] hub) and a post-reframe deep lint (100 pages, 3 agents) confirmed excellent structural health with the reframe propagated cleanly. The detailed Stage record below is an accurate ingest history. [2026-06-14] Pilot design crystallized into the new analysis page Thesis Objectives and Grant Pitch (strategic/objectives layer; complements the Rev 4 Research Plan - DN B Cell Expansion in Dengue) across a multi-turn brainstorm: gating strategies → fixed-panel resolution ceiling → cross-sectional d5–8 pilot (DF=8/DHF=11, target ≥10–15/arm) → grant pitch. Now has an antibody layer (ANA + FRNT×4 + IgG/IgM) making both faces of the bridge thesis measurable in-pilot; O1 = cells→ANA correlation as the novel primary. See new Watch Items.

• Scope: Atypical / age-associated B cells (DN/DN2/ABC/T-bet⁺) and plasmablasts in dengue infection — their phenotype, the extrafollicular pathway that generates them, and their antibody output (autoreactive + neutralizing) — characterised primarily through flow cytometry (conventional, spectral, high-dimensional). EF is now framed as the generating pathway under the atypical-cell spine, not the wiki’s organizing frame. The autoantibody / neutralizing-antibody association is synthesized cross-wiki in ../bridge-wiki/. (Also includes isotype switching and the relationship to clinical outcomes — severity, primary vs. secondary infection, serotype.)
• Stage: Twenty sources ingested (11 comparative + 8 dengue-specific + 1 dengue commentary). Wiki has 99 pages across 20 sources (Lamprinou2026 ingested 2026-06-14). New analysis: DN2 Gating Strategy — council-reviewed 11-color gating strategy for isolating DN/DN2-phenotype B cells from dengue PBMCs, directly comparable to Ansari2025. Council endorsed with 4 modifications (generous FSC/SSC gate, polygon gates, CD11c-PE FMO mandatory, “DN2-phenotype” terminology). William2002 (Science) now ingested as the foundational murine proof of extrafollicular SHM — previously the most-cited external reference in the wiki (4 bare citations converted to wikilinks). Establishes EF SHM at GC-comparable rates (~0.3 mut/gene/generation), EF tolerance escape mechanism, and TLR9 co-stimulation as EF driver (the mechanistic precedent for TLR7 in dengue). New method page: Immunohistochemistry. Ansari2025 is the landmark paper — first phenotypic evidence consistent with EF B cell activation in dengue (tissue confirmation lacking), identifying Tph→IL-21→memory B cell→plasmablast as the dominant B cell help axis, with CD21⁻CD11c⁺ (DN2-phenotype) B cells expanded in acute dengue. Council review completed (2026-05-14): IL-21 coculture rated STRONG; severity association downgraded (day-of-sampling confounder); “Tph” identity debatable (Th1 signature, not canonical Tph); scRNA-seq is 3 patients (hypothesis-generating only). The comparative framework (SLE + COVID-19) is extended to dengue but not fully validated (DN2 identity unconfirmed, naive vs. memory tension unresolved). A detailed wet-lab research plan exists (analyses/Research Plan - DN B Cell Expansion in Dengue.md) — Revision 4 (2026-05-24), now incorporating Kaneko2020 (tissue-level GC loss), William2002 (EF SHM proof), Sutton2021 (alternative lineage framework/gating caveat), Bhattacharya2016 (tissue-retained PBs), Priyamvada2016, Singh2026, and GarciaBates2013. Systematic “DN2-phenotype” terminology applied throughout. H4 softened from precursor–product to co-variation model per Sutton2021. New Follow-Up Study 8 (CD11c-primary gating reanalysis). 12 limitations (up from 10). Follow-Up Study 4 (BCR sequencing of sorted DN2-phenotype cells) carries specific falsifiable predictions from the molecular evidence. GodoyLozano2016 provides the first direct SHM measurement in dengue IgG B cells — paradoxically low SHM in acute phase, lower in secondary than primary, lower in DWS+ than DWS−. This is the strongest molecular evidence for EF pathway operation in dengue, with IGHV1-2/1-69 bias and convergent CDRH3s at 52% prevalence. The dual EF+GC model is now supported by both cellular (Ansari2025) and molecular (GodoyLozano2016, Parameswaran2013) evidence. Priyamvada2016 provides the sorted PB BCR sequencing that was the key remaining gap — but with opposite result to expected: HIGH SHM (mean 18.1 VH mutations) in sorted secondary DHF PBs, supporting memory B cell origin. This creates a dual-pathway model: memory-derived high-SHM PBs (Priyamvada2016) + de novo EF-derived low-SHM PBs (GodoyLozano2016). Also establishes two new concept pages: Original Antigenic Sin (OAS in 2/4 patients — DENV1 preferentially neutralised over infecting DENV2) and Antibody-Dependent Enhancement (45/53 mAbs ADE-competent regardless of neutralisation). Key remaining gap: no study has simultaneously sorted PBs from primary vs. secondary dengue for SHM comparison — this would test whether primary PBs are low-SHM (de novo EF) while secondary PBs are high-SHM (memory recall). Sutton2021 (Cell Reports) provides the transcriptomic framework for recontextualising DN2/atypical B cells as part of an “alternative lineage” distinct from classical memory. CITE-seq demonstrates CD21⁻CD27⁻ gating captures only 44.7% of transcriptomic atBCs — CD11c is the best single surface marker. ~20% of B cells in healthy donors belong to the alternative lineage. Critically, no atBC cluster upregulates PC maintenance genes (XBP1, IRF4, PRDM1), arguing against the EF pre-PB model in healthy/infection contexts — but Sutton reconciles this as context-dependent (SLE TLR7 may drive PC fate). MBC1 cluster provides transcriptomic evidence for “memory DN2” (partially confirming Sanz2025/Faliti2024). New method page: CITE-seq. Kaneko2020 (Cell, 2020) provides the tissue-level histopathological foundation for the EF dominance model in acute viral infection. Post-mortem COVID-19 tissue (n=11 + controls) shows complete GC absence: Bcl-6⁺ GC B cells and Bcl-6⁺ GC-TFH both absent, but AID⁺ B cells preserved — establishing that SHM/CSR enzymatic machinery operates outside GCs. TNF-α accumulation proposed as the TFH differentiation block mechanism. TH1 (T-bet⁺) CD4⁺ expansion in tissue parallels dengue Tph but with different functional implications (GC suppression vs. B cell help). Peripheral blood (n=68, BD Symphony 13-color panel) confirms DN/aN/PB expansion with SARS-CoV-2 RBD specificity by dual-fluorophore probes. Was previously cited as bare external reference in 4 wiki pages; all converted to proper wikilinks. New entity pages: Bcl-6, AID, TNF-alpha, ICOS. New method page: Multi-color Immunofluorescence. Key dengue implication: does severe dengue produce enough TNF-α to disrupt GC TFH differentiation? If so, this would mechanistically explain the low-SHM IgG (GodoyLozano2016) and EF dominance. Lamprinou2026 (Frontiers in Aging, opinion piece — no original data, self-cited taxonomy) is the wiki’s first source dedicated to the ABC ↔ DN identity question the Atypical B Cell umbrella was built to map. Establishes ABC as a heterogeneous superset (CD27⁺ + IgD⁺ + predominantly IgD⁻CD27⁻) whose IgD⁻CD27⁻ fraction maps to DN2 — partial, asymmetric, context-dependent overlap, and even there ABC ≠ DN2 transcriptomically (Maul 2021). Mechanistic split: IL-21→CD11c, IFN-γ→T-bet. Introduces a four-subset DN taxonomy (adds DN4, allergy-associated, CXCR5⁺) vs. the wiki’s default three; logged as nomenclature drift. New entity page: Age-Associated B Cell. No dengue data — comparative/conceptual anchor only. This refines but does not overturn the spine reframe; the umbrella synonymy map’s “ABC ≈ DN2” row may warrant revision to an explicit asymmetric-overlap (see Watch Items). Next priority: papers with T-bet staining to confirm DN2 identity; studies on OAS across serotype combinations; ADE functional studies.
• Open questions to keep in mind as ingests proceed:
  • Are the CD21⁻CD11c⁺ B cells in dengue truly DN2 (T-bet⁺, FCRL5⁺, CXCR5⁻)? Needs intracellular T-bet staining.
  • What is the SHM distribution of sorted dengue plasmablasts? GodoyLozano2016 shows globally low SHM in IgG B cells during acute phase (Monte Carlo simulation supports ASC dominance), but formal sorted PB BCR sequencing would conclusively demonstrate germline (EF) vs. mutated (GC) origin.
  • Does the Tph→IL-21 axis operate differently in primary vs. secondary dengue?
  • Is the Tph pathway the source of ADE-competent cross-reactive IgG in secondary dengue?
  • Does IL-21 blockade reduce severity-associated (non-neutralizing) antibodies while preserving neutralizing titers?
  • What is the functional role of cytotoxic (GZMB⁺HOPX⁺) Tph cells?
  • Do EF and GC responses truly operate concurrently in dengue, or does the balance shift with disease stage?

Queue

Next session first step: Resume paper ingest queue (balakrishnan2011, kwissa2014, woda2016, zompi2012).

Papers waiting to be ingested (add new entries at the top):

• morra2018.pdf ✅ ingested 2026-06-29 (second severity-axis source; within-scheme definitional heterogeneity)
• balakrishnan2011.pdf
• kwissa2014.pdf
• Bhattacharya2016.pdf ✅ ingested 2026-05-19
• woda2016.pdf
• william2002.pdf ✅ ingested 2026-05-18
• zompi2012.pdf
• Sutton2021.pdf ✅ ingested 2026-05-22
• Kaneko2020.pdf ✅ ingested 2026-05-22

Decisions

Structural and workflow decisions with rationale. Append-only.

[2026-05-02] Naming conventions established (first ingest)

Decision: Entity pages use singular descriptive names (Double-Negative B Cell, Plasmablast, CD27). Source short titles follow AuthorYear - Short Descriptive Title format (e.g., Wei2007 - DN Memory B Cells in SLE). Concept and method pages use plain descriptive names (Extrafollicular Response, Conventional Flow Cytometry). Why: First ingest sets the template; retroactive renaming across many pages is costly. How to apply: Follow these conventions for all future ingests. If a new paper uses a term that conflicts with an existing page name, prefer the existing page name and add the alternative term in the Overview section.

[2026-05-02] DN B cells vs. atypical B cells — treated as overlapping, not identical

Decision: Double-Negative B Cell (IgD⁻CD27⁻) is the primary entity page for this population, per curator direction. It is noted as overlapping with, but not identical to, “atypical B cells,” “age-associated B cells (ABCs),” and “T-bet⁺ B cells” used in later literature. These terms will not get separate entity pages until a paper provides a direct phenotypic comparison justifying the distinction. Why: Curator specified Double-Negative B Cell as the entity name; the field uses multiple overlapping terms for similar populations. Premature splitting would fragment the evidence base. How to apply: When a paper uses “atypical B cell” or “ABC” terminology, cross-reference to Double-Negative B Cell and note any phenotypic differences (e.g., T-bet expression, FcRL5, CD11c) in the entity page.

[2026-05-02] Bm Classification → method page; 9G4 tracking → out of scope for dengue

Decision: During Anolik2004 ingest, curator directed: (1) create a method page for the Bm1–Bm5 classification framework (done: methods/Bm Classification.md); (2) skip the 9G4 VH4.34 antiidiotype tracking method as out of scope — it is SLE-specific autoreactivity tracking with no direct dengue application. Why: Bm1–Bm5 is a foundational gating framework used across multiple B cell biology papers and likely appears in dengue flow cytometry studies; it warrants its own page. 9G4 is a lupus-specific reagent; tracking it would bloat the wiki without dengue relevance. How to apply: If a future paper uses 9G4 or VH4.34 tracking in a dengue context, reconsider. Otherwise, treat 9G4 results as background context only (captured in source pages) rather than creating a method page.

[2026-05-08] Insertion-order guard added to Ingest workflow; sub-agent lint codified

Decision: Two CLAUDE.md workflow updates: (1) Ingest steps 5–7 now include an explicit insertion-order guard — new Key Points must be inserted at the end of ## Key Points from Literature, not appended at end-of-file. (2) Lint workflow now recommends parallel sub-agents for deep lints. Why: The 2026-05-08 deep lint found 19 pages with content displaced after ## Related Pages due to end-of-file appending during ingests. Sub-agent parallelism was validated as significantly more efficient for scanning 79 pages. How to apply: During every ingest, when updating entity/concept/method pages, locate ## Contradictions & Debates or ## Related Pages and insert new bullets immediately before it. During deep lints, spawn parallel agents per folder batch.

[2026-05-11] Web deployment live — efb-dengue-wiki.pages.dev

Decision: Site deployed to Cloudflare Pages via GitHub repo OsandaC/efb-dengue-wiki (branch main). Quartz v4 static site in efbwebshare/ folder (sibling of efb-dengue-wiki/). sync-and-build.ps1 syncs efb-dengue-wiki/wiki/ → efbwebshare/content/, builds, commits, and pushes. Cloudflare auto-deploys on push. Update Web workflow added to CLAUDE.md. Why: Curator ready to share the EFB wiki publicly, replicating the proven dengue-wiki deployment pattern. How to apply: When curator says “update web”, run sync-and-build.ps1 from efbwebshare/. After Dependabot PRs on GitHub, run npm install in efbwebshare/ and push updated package-lock.json before next deploy.

[2026-05-14] “Summon the Council” workflow added

Decision: New CLAUDE.md workflow — multi-agent critical review panel. 4 default roles (Methodology Critic, Claims Validator, Contextual Critic, Strengths Advocate) + Council Head. Curator can add/replace/reduce roles. Output goes to Claude-council/ as two files per paper (raw minutes + final report). Reports only — no wiki modifications unless curator directs. PDF reader required; if unavailable, halts and asks curator before falling back to wiki source page. Why: Curator wants structured multi-perspective critical review of papers. First council (Ansari2025) demonstrated the value of parallel specialist agents catching different issues. Codified to make repeatable. How to apply: When curator says “summon the council” or “council review” + paper name, follow the 7-step workflow in CLAUDE.md §Workflows → Summon the Council. Check PDF reader availability first. Deploy agents in parallel. Do not modify wiki pages based on council findings unless explicitly instructed.

[2026-05-14] Poppler fallback path added to Council workflow

Decision: CLAUDE.md §Workflows → Summon the Council → Step 2 now includes a hardcoded fallback path for pdftoppm.exe at C:\Users\user\AppData\Local\Microsoft\WinGet\Packages\...\poppler-25.07.0\Library\bin\pdftoppm.exe, used when pdftoppm is not on the system PATH. Why: Poppler installed via WinGet but not added to PATH. Without the fallback, council workflow would halt unnecessarily. How to apply: If poppler is updated to a newer version or moved, update the path in CLAUDE.md.

[2026-05-24] Schema refactor — lazy-load council, rename governance, trim fat

Decision: Three-part CLAUDE.md refactor: (1) Council workflow (106 lines) extracted to lazy-loaded CLAUDE_COUNCIL.md — read only when curator summons the council; (2) CLAUDE_UPDATE.md renamed to CLAUDE_GOVERNANCE.md — clearer name, baseline improvements table moved to log.md; (3) New Axis / Remove Axis stubs replaced with pointer to governance file, eliminating duplicate source-of-truth. Net: 114 lines (~2,400 tokens) removed from per-session context load. Why: Curator-directed efficiency refactor. Council protocol loaded every session (~1-in-10 usage). Axis stubs duplicated governance file. Baseline table was historical commentary, not operational. How to apply: When council is summoned, read CLAUDE_COUNCIL.md (not CLAUDE.md — it only has a stub). When axis changes are requested, read CLAUDE_GOVERNANCE.md §Change Types. The governance file is still the authority for all schema changes.

[2026-05-02] Wiki initialised

Decision: Scaffold created as a sibling to the dengue-wiki, focused on extrafollicular B cell dynamics in dengue via flow cytometry. Adapted from dengue-wiki schema with three differences: (1) geography/ axis dropped (clinical cohort source captured in source page metadata only); (2) Update Web workflow removed (web deployment deferred); (3) Domain Context rewritten for B cell / cytometry scope. Why: Curator wants a focused, blank-slate wiki for a tighter research question, leveraging the proven schema/workflow design from the dengue wiki. How to apply: Treat geography as out-of-scope unless the curator re-introduces it. If a paper’s cohort location is relevant, capture it in the source page’s “Setting” field — do not create geography pages.

[2026-06-13] New high-level direction + three-wiki architecture

Decision: New direction — “plasmablasts and atypical/age-associated B cells and their association with autoantibodies and neutralizing antibodies in dengue.” Resolved into three sibling wikis: dengue-wiki/ (antibodies/ANA/autoimmunity — canonical), efb-dengue-wiki/ (this — cells/EF/atypical/plasmablasts — canonical), bridge-wiki/ (new synthesis layer). Seam: option (c), a standalone bridge artifact with its own state+log; no merge — both parents exceed comfortable session-context limits. Spine reframes toward atypical/age-associated B cells + plasmablasts; EF demoted to a pathway under them. Unifying thesis: one low-fidelity antibody property (cross-reactive, polyreactive, near-germline) with two faces (autoreactivity via mimicry + non-neutralization/ADE); anti-NS1/anti-prM are the linchpins. Novel/unbuilt arm = cells→autoantibodies; neut/ADE/OAS already double-covered in both parents (reconcile, not build). Why: The autoantibody literature is already fully built in dengue-wiki (~45 sources, ANA/mimicry); this wiki owns the cells. The genuine gap is the cellular origin of dengue autoantibodies — a bridge between two existing wikis. Curator-directed; advisor-reviewed. How to apply: Build cells→autoantibody synthesis in bridge-wiki/, citing parent pages with efb-wiki: / dengue-wiki: prefixes; do not re-ingest. Treat the cells→autoantibody link as an SLE-imported hypothesis, not dengue fact. Before reframing efb structure (Atypical B Cell page; Current Focus/CLAUDE.md spine), get curator go-ahead.

[2026-06-13] Spine reframe executed — Atypical B Cell umbrella created (supersedes [2026-05-02] atypical-split decision)

Decision: Created the [[Atypical B Cell]] umbrella/hub entity page and reframed the spine toward atypical/age-associated B cells + plasmablasts (EF demoted to a generating pathway). This supersedes the [2026-05-02] “DN vs. atypical — no separate atypical/ABC pages until a paper provides direct phenotypic comparison” decision: the umbrella is now justified not by a new phenotyping paper but by the elevation of this cluster to the wiki’s organizing spine. DN nomenclature remains the precise classification (the umbrella foregrounds the Sanz2025 argument that “atypical” conflates ≥5 populations); the Atypical B Cell page is a synonymy-map hub routing to the DN sub-pages, not a competing content page. Why: Curator-directed new direction; the atypical/plasmablast cluster is the cellular half of the cross-wiki bridge to autoantibodies/neutralizing antibodies. How to apply: Treat Atypical B Cell as the umbrella entry; keep detailed evidence on the DN/DN2/DN3/aN sub-pages. The CLAUDE.md identity wording remains gated on the curator’s term choice (see Watch Items).

[2026-06-27] Voice (TTS) Mode documented as a CLAUDE.md workflow; server lifecycle coupled to the mode

Decision: Added ### Voice (TTS) Mode to CLAUDE.md §Workflows (between Update Web and Lint). /tts-on now starts the shared Kokoro TTS server (127.0.0.1:8880, GPU) detached and /tts-off stops it, so the server is only resident while voice is on. New scripts start_kokoro_detached.ps1 / stop_kokoro.ps1 under ~/.claude/voice-mcp/; /tts-on and /tts-off command files updated to call them. Applied to CLAUDE.md §Workflows. 0 wiki pages migrated. Why: Curator did not want a GPU TTS server running constantly; wanted a start/stop trigger tied to the voice toggle. These are global Claude Code commands (not wiki content); documented here for discoverability only, mirroring the Update Web precedent. How to apply: When voice is needed, /tts-on (idempotent; aborts if the server won’t pass /health); when done, /tts-off (frees the GPU). The server is shared with live-narration — skip the stop step if voice should be off but the server kept up. All scripts/commands live under ~/.claude/, outside this repo, so they are not covered by wiki git snapshots.

[2026-06-27] Switched Memory (sM) added as a population of interest; expanded gating tree reconciled

Decision: Created [[Switched Memory B Cell]] entity page (IgD⁻CD27⁺, GC-derived memory) as the germinal-center comparator to the DN/DN2 EF spine, and folded the curator’s expanded full-B-cell gating tree into [[DN2 Gating Strategy]] — an sM-isolation subsection (resting CD21⁺ / activated CD21⁻ split) plus a 4-point reconciliation of overlaps in the pasted tree. No new paper ingest (synthesized from existing sources); index Entities 47→48. Why: Curator is expanding the 11-color panel beyond DN-only to a full B-cell immunophenotype and adding sM as a gate. sM is the highest-value addition because it is the matched GC control for the EF/DN story — sM (IgD⁻CD27⁺) and DN (IgD⁻CD27⁻) sit side by side in the IgD/CD27 plot, differing only by CD27. How to apply: Report sM and DN as a matched GC-vs-EF pair. Gate IgD⁻ first before any resting/activated memory split so it is switched-specific. Use one label (“DN2-phenotype”) for CD21⁻CD11c⁺ within DN; include the CD21⁻CD11c⁻ (DN3-like) quadrant so DN subsets sum to 100%. The sM↔DN boundary is CD27-shedding-sensitive in dengue’s high-TNF setting — carry as a caveat.

Watch Items

Tracked issues, gaps, and follow-ups. Resolve and remove as addressed.

• [2026-06-25] PhD upgrade meeting — Prof. Rukie and Prof. Neelika (2026-06-26) — DONE, WENT WELL. Outcome (curator report 2026-06-27): meeting went well; buy-in secured on the initial verbal talk alone — the LLM/wiki demo was not needed in the pitch. Direction endorsed; cleared to proceed. Next phase per curator: add more papers + fine-tune the research directions (see new Watch Item below). The pre-meeting plan/fact-check history (n=19 fix, Q&A rehearsal, wiki-first demo order) is preserved in wiki/log.md (2026-06-25/26 entries) and the plan file ~/.claude/plans/pure-zooming-adleman.md.

  • [2026-06-25 update] Prep finalized. Fact-checked the plan against sources + the Ansari2025 council report: fixed the GodoyLozano2016 n=175→n=19 error (in 2 places) and the stale “99 pages”→“100+ / 20 papers” claim; added an O1-is-novel line, a 7th Q&A row (primary-endpoint defence), and sharpened the 4 explicit asks into spoken questions; ran a cold rehearsal of all 7 hard questions. Plan is meeting-ready. Q1 (Tph identity) + Q2 (SHM paradox) are the two answers that can spiral under follow-up — safety-pivot on both is to redirect from the weak link to what the pilot measures (cells→ANA). Fact-check details in wiki/log.md (2026-06-25 prep entry).
  • [2026-06-26 update — morning-of reframe] This is the PRELIM meeting → wiki-first, not pitch-first. Curator clarified at the door: Prof. Rukie is most likely interested in the wiki as a tool/method, and this is the first introduction of the wiki + Claude to her. Lead with: one-breath intro (living lit-review KB, 20 papers / 100+ interlinked pages, Claude-maintained); three rigor points (evidence weighting w/ study-type+n inline; flags contradictions & downgrades claims — Ansari severity downgrade, Tph doubt; the Council adversarial layer); demo order by impact (Obsidian graph → Ansari2025 source page → live site efb-dengue-wiki.pages.dev). Introduce Claude plainly (“Claude summarizes/cross-references/maintains; I curate and direct”); if asked about hallucination, answer that everything traces to a source page and is fact-checked pre-meeting. The PhD pitch (O1, pilot, asks) stays in the back pocket unless she steers there. Next sign-in is live, in front of Prof. Rukie. Resolve this whole item after the meeting.

• [2026-06-27] Post-upgrade next phase — add papers + fine-tune directions. Upgrade endorsed; now expanding the evidence base and refining research directions. Queue still holds 4 un-ingested papers (balakrishnan2011, kwissa2014, woda2016, zompi2012 — all in raw/). Curator will also add new papers and may adjust the thesis directions (O1 cells→ANA primary; pilot design). Pending curator input: (a) which new papers to add and on what sub-topic; (b) what direction changes (if any) came out of the meeting. Until specified, default next action = resume the existing ingest queue.

• [2026-06-16] CD19+CD66b+ population — gating decision pending (RESUME HERE). During dump-gate validation on Specimen_001_HT 82_002.fcs, found 5,789 CD19+CD66b+ singlet events (1.70% of live leukocytes; 19.5% of the current B cell denominator of 29,665). These are confirmed not doublets (FSC-A vs FSC-H is tight diagonal; two singlet gates already in hierarchy). Spillover ruled out: the brightest CD19 events (main B cell cluster at CD1910⁴) are the most CD66b-negative — PerCP-Cy5-5→PE-Cy7 overcompensation would produce the opposite pattern. Conclusion: real CD19+CD66b+ single cells being excluded by the dump gate. Decision pending — two options: (a) widen/lower the CD66b dump boundary to recover them, accepting some granulocyte contamination risk; (b) keep gate as-is and note as a formal limitation (B cell denominator excludes CD66b+ B cells, likely enriched for activated/plasmablast phenotypes relevant to dengue). If all 5,789 recovered: true B cell N = 35,454; DN% = 2,640/35,454 = 7.45% (vs current 8.90%). Next step: decide option (a) vs (b), then characterise the CD19+CD66b+ population on other markers (IgD, CD27, CD21, CD11c) to assess whether it is plasmablast-enriched.

• [2026-06-14] Age/sex balanced recruitment — first-order, act now. Atypical/age-associated B cells are age-defined and female-biased (X-linked TLR7); current arms are imbalanced (DHF male-skewed/older; DF female-skewed/younger). At n≈10–15/arm, multivariable adjustment is weak → recruit remaining DF/DHF (and the comparator arm) to balance sex and age across arms by design. Tracked on Thesis Objectives and Grant Pitch.

• [2026-06-14] Severity-scheme decision pending (curator’s call). Recommendation on the page: pre-register WHO-2009 binary as primary + WHO-1997 leak-defined as declared sensitivity analysis; 2 strata at this n; classify on full clinical course. [2026-06-29 update] Evidentiary anchor now in-wiki: Dengue Severity Classification (from Narvaez2011) quantifies the trade-off — WHO-2009 sens 92.1% vs WHO-1997 39.0% for intervention-level disease; κ=0.25 between schemes; the DENV-2 severity signal is scheme-dependent (present under WHO-1997, absent under WHO-2009). Decision still curator’s call. [2026-06-29 — Morra2018 adds a finer layer] Beyond which scheme, the pilot should pre-register the operational sign-definitions (the actual cutoffs for warning signs / severe dengue), not just the scheme name: Morra2018 - Defining Warning Signs and Severe Dengue shows “WHO-2009” alone under-determines the case definition (only 2 of 16 signs reach a consensus definition; “shock” defined 23 ways), so a scheme label without fixed cutoffs leaves the severity strata non-reproducible and non-poolable with the literature. (Closes the Morra source-page Q4.)

• [2026-06-29] Deferred: backlink the severity corpus to Dengue Severity Classification. The Narvaez2011 ingest created the concept hub but only curated-linked 5 high-value pages (Research Plan, Thesis Objectives, GarciaBates2013, GodoyLozano2016, Ansari2025) via lightweight inline see-references; the concept page’s ## Sources lists only Narvaez2011. ~35 other pages mention DF/DHF/DSS/Severe Dengue/Warning Signs with no link. Next lint: expand the concept page’s ## Sources to the severity-stratified dengue sources and add proper backlinks (Wrammert2012, Priyamvada2016, Plasmablast, Singh2026, etc.), tagging each with the scheme it used (WHO-1997 / WHO-2009 / national Brazilian DF-DFC). Deferred deliberately to avoid a propagation balloon during the ingest. [2026-06-29 update] Also fold in Morra2018 - Defining Warning Signs and Severe Dengue (the within-scheme source): the concept page’s ## Sources now lists Narvaez2011 + Morra2018, and the Morra ingest added only 2 curated source-page links (GarciaBates2013, GodoyLozano2016). When the backlink lint runs, additionally note — where the source reports it — the operational sign-definitions a study used, not just the scheme name (Morra’s point is that the scheme label alone under-determines the case definition).

• [2026-06-14] HC/comparator + convalescent arm status unknown. Is there an age/sex-matched healthy/seropositive comparator? Any convalescent draw feasible on a subset (would unlock ANA resolution + kinetics — major strengthener)? Single acute timepoint currently limits the ANA-specificity argument to the within-cohort correlation.

• [2026-06-14] Serology + serotype operational items. LFA on all → capture-ELISA IgM/IgG ratio on a calibration subset → compute LFA↔ELISA κ (pre-set threshold) to decide if LFA suffices broadly. Attempt RT-PCR/NS1 on the d5 subset to anchor the epidemiological DENV2 assumption (constrains OAS interpretation only).

• [2026-06-14] Research Plan Rev 5 recommended (not yet done). Fold the antibody assays (ANA, FRNT×4, IgG/IgM), the d5–8 window (vs current 5–9), realistic n, age/sex balancing, and the new intracellular panel into Research Plan - DN B Cell Expansion in Dengue when the curator is ready. Rev 5’s primary panel should be B Cell Panel Variant 1 Panel 4 (13-color single-tube, intracellular T-bet + CXCR5), replacing the surface-only 11-color from DN2 Gating Strategy — the lab now has intracellular-staining capability.

• [2026-06-14] Panel decision: single-tube vs suite (curator’s call). B Cell Panel Variant 1 recommends Panel 4 (single tube) for all patients; Panels 1–3 are a multi-tube suite gated by cell yield/budget. Deciding fact = cells per fresh PBMC draw + per-tube cost (unknown). If multi-tube is feasible on a subset, Panel 2 (ASC output) > Panel 3. Pre-bench to-dos: confirm chosen conjugates exist + pass a spillover-spreading matrix before ordering; validate T-bet-APC + Ki-67-BV711 survive the lab’s TF perm buffer.

• [2026-06-27] Confirm the expanded full-B-cell gating tree + fold into the primary panel docs. The curator’s new tree extends beyond DN-only: plasmablasts (now kept as a population, not just an exclusion gate), CD27⁺ memory resting/activated subsets (CD27⁺CD38⁻CD24⁺CD21±), naïve, sM (IgD⁻CD27⁺), and the DN 2×2. [[DN2 Gating Strategy]] now documents the sM gate + 4 reconciliation points, and (2026-06-27) carries a consolidated “Canonical Gating Tree (Start → All Subpopulations)” subsection — the full start→finish hierarchy with a 9-population terminal checklist (incl. the kept-plasmablast branch and the memory resting/activated split). The canonical tree is therefore confirmed within the gating-strategy page; what remains is folding it into [[B Cell Panel Variant 1]] (Panel 4) and Research Plan Rev 5 (pairs with the existing Rev 5 watch item). Open design point already baked into the tree: gate IgD⁻ first so the resting/activated memory split is switched-specific (else it overlaps the sM quadrant).

• [2026-06-14] Missing entity page: Ki-67. Referenced inline across the wiki (Plasmablast “Ki67⁺”; DN2 Gating Strategy missing-marker list; now B Cell Panel Variant 1) but has no entity page — a [[Ki-67]] link was dropped from the new panel page to avoid a dangling reference. Create a proliferation-marker entity page on the next relevant ingest (pairs with CD71).

• [2026-06-14] Compensation/FMO walkthrough — DONE. Created wiki/methods/Compensation and FMO Controls.md (general principles + curator’s 11-color DN2 panel worked example). Compensation matrix validated as sound (cond=6.22, det=0.907); CD21→CD11c spillover (~20%, the DN2-axis pair) confirmed clean. FMOs for CD11c/CD21/CD27/IgD revealed the working DN cutoffs (IgD<0.8, CD27<1.0) were undercounts — FMO 99th-pct boundaries (IgD<1.98, CD27<1.76) are now adopted, quadrupling DN from 1.99%→8.90% of B cells; DN2 (CD21<0.69 & CD11c>0.72 within DN) = 7.99% of DN.

• [2026-06-14 → RESOLVED 2026-06-27] Naming drift in DN2 Gating Strategy.md panel table. Corrected the panel table and the gating-step axis labels: RB705→PerCP-Cy5-5 (CD19), eFluor506→AmCyan (L/D), APC-Fire750→APC-H7 (CD45). Done during the sM-gate edit pass.

• [2026-06-14 → RESOLVED 2026-06-27] DN2 Gating Strategy.md described input as “PBMCs”; actual prep is whole-blood RBC-lysate. Research Question now reads “whole-blood leukocytes (RBC-lysed whole blood, not Ficoll-separated PBMCs)”; the incidental “adult dengue PBMCs” in the transitional-rarity note changed to “adult dengue whole blood.” Done during the sM-gate edit pass.

• [2026-06-27] Residual fluorochrome-naming drift in [[Research Plan - DN B Cell Expansion in Dengue]]. That page’s panel table, comp-control list, and gating steps still use the pre-pilot names RB705 (CD19), eFluor506 (L/D), APC-Fire750 (CD45) (~11 occurrences). Now that [[DN2 Gating Strategy]] is corrected, the Research Plan is the remaining page carrying the old names. Reconcile during the Rev 5 pass (pairs with the Rev 5 watch item) — deliberately not bundled into the sM edit to avoid a large unreviewed change to a primary planning doc.

• [2026-06-14] AmCyan-A (L/D) single-stain control format unconfirmed. Compensation and FMO Controls says the comp matrix came from “bead-based single-stain controls” but doesn’t specify the viability-channel control’s format. AmCyan-A (amine-reactive) doesn’t bind standard comp beads — if a standard bead was used for this channel, that coefficient is likely near-zero/unreliable and the live/dead gate (Step 1, DN2 Gating Strategy) would be effectively uncompensated. Confirm against run records before the next staining batch; flagged in DN2 Panel - Staining, Compensation, and Gating Protocol.

• [2026-06-14] Bead-level validation of 3 spillover coefficients >100%. BV711→AF700/BV786 (~119–148%) and PE-Cy7→CD19 (~102%) were shown harmless for DN/DN2 gating (their source channels are dump channels removed pre-B-cell-gate) but not validated at the bead level. Non-blocking; revisit if single-stain bead FCS files become available.

• [2026-06-14] CD11c FMO threshold (0.72) is low-precision (~4 tail events of 387). DN2% (7.99% of DN) should be treated as approximate until a larger CD11c-PE FMO acquisition or pooled-sample validation is run.

• [2026-06-13] [[Atypical B Cell]] umbrella page — DONE. Created as a synonymy-map hub; Double-Negative B Cell reframed to defer to it; index updated (Entities 45→46, pages 96→97). Superseded the [2026-05-02] no-split decision.

• [2026-06-13] efb spine reframe (state.md/pages) — DONE. Current Focus Scope reframed to atypical cells + plasmablasts (EF = generating pathway).

• [2026-06-13] CLAUDE.md identity rewrite — DONE (hybrid “Atypical (DN)” term). Curator chose the hybrid. CLAUDE.md H1, mission, and Domain Context intro rewritten per CLAUDE_GOVERNANCE.md (minimal diffs); reframe log entry updated. The spine reframe is now complete across pages, index, state, and the instruction set.

• [2026-06-13] Cross-wiki ADE/OAS reconciliation. ADE and OAS are concept pages in BOTH this wiki and dengue-wiki with different source sets (ADE 5 here / 12 there; OAS 7 here / 3 there). Latent contradiction. When the bridge synthesis touches this arm: decide canonical owner and diff the two pages. Also tracked in bridge-wiki/state.md.

• [2026-06-14] Revise the Atypical B Cell umbrella synonymy map: “ABC ≈ DN2” → explicit asymmetric overlap. Lamprinou2026 (and Maul2021) establish that ABC is a heterogeneous superset intersecting DN only at the IgD⁻CD27⁻/DN2 node, and that even that intersection is phenotypic not transcriptomic. The umbrella’s ABC row was softened during this ingest, but a clearer asymmetric-overlap diagram (ABC ⊅ DN, DN ⊅ ABC) would prevent silent transfer of findings between “ABC-gated” and “DN-gated” studies. Flagged in Notable Findings #16.

• [2026-06-14] DN4 in dengue + CXCR5⁺-DN gating blind spot. Lamprinou2026 adds DN4 (CXCR5⁺CD11c⁻T-bet⁻, allergy-associated) to a four-subset DN scheme. CXCR5⁻-focused EF gating (Ansari2025-style) systematically discards the CXCR5⁺ DN subsets (DN1, DN4). Does DN4 exist in dengue, and does the standard EF gate miss a relevant compartment? Needs a dengue panel with CXCR5 within the IgD⁻CD27⁻ gate. (Overlaps existing DN1/DN2-ratio and CXCR5-staining watch items.)

• [2026-06-14] Is the IgD⁻CD27⁻ ABC subset transcriptomically identical to DN2? Maul2021 (via Lamprinou2026) shows ABC ≠ DN2 across the whole superset, but whether the cytokine/chemokine distinction persists within the IgD⁻CD27⁻ fraction is untested. Requires paired transcriptomics of sorted IgD⁻CD27⁻ ABCs vs. DN2.

• [2026-06-14] Species discordance in ABC depletion sensitivity. Murine ABCs resist anti-CD20/anti-BLyS depletion (Knox 2025) while human SLE ABCs are sensitive (Ramsköld 2018; Faustini 2022) — flagged on Age-Associated B Cell and CD20 pages. Species difference vs. disease-stage/context difference is unresolved; relevant to whether B-cell-depletion findings transfer across the comparative benchmark species.

• [2026-06-14 deep lint] Post-reframe + Lamprinou health check (100 pages, 3 agents) — DONE. 0 HIGH; wiki in excellent structural health, reframe propagated cleanly (no stale flat ABC=DN2). Fixed 8: 6 source→hub back-links (Sanz2025/Sutton2021→both hubs; Jenks2018/Woodruff2020/Ansari2025/Singh2026→Atypical B Cell), EF Response Overview reframe-framing, T-bet Overview equivalence soften. Curator Highlights refreshed (2, unchanged).

• [2026-06-14 deep lint] Minor: IgA and IRF4 each list a source under ## Sources not cited in any Key Points bullet. Frontmatter counts still match (not a defect). Either add a citing bullet or drop the uncited source on next edit of these pages. Low priority.

• [2026-06-14 deep lint] External Citation Audit is stale. 2026-05-08 snapshot built against 8 sources; wiki now has 20. Lamprinou2026’s external cites (Cancro 2020, Tangye 2023, Maul 2021, Ambegaonkar 2022, Naradikian 2016, etc.) are uncatalogued. Refresh if/when the curator resumes the external-citation review (folds into the existing [2026-05-08] External Citation Audit watch item).

• T-bet / FcRL5 / CD11c / CXCR5 entity pages thickened. Originally single-source (Jenks2018); all now at 2 sources after Sanz2025 ingest.

• Plasmablast page — no longer a thin stub. Substantially populated from Anolik2004 (expansion data, CD20⁻ phenotype, short-lived kinetics, long-lived vs. short-lived dichotomy). Still requires a dengue-specific plasmablast paper for dengue-specific kinetics and marker panel details.

• [2026-05-02 lint] Frontmatter source count mismatches fixed. CD10, IgG, IgM, IgA, FACS Sorting, Class Switch Recombination all had sources: 1 in frontmatter but 2 sources listed in body. Corrected. Conventional Flow Cytometry duplicate “Related Pages” section removed.

• [2026-05-02 lint, updated 2026-05-22] Remaining thin pages (single source). Entities (10): B220, CD23, CD71, ATF3, EGR, HOPX, Peripheral Helper T Cell, TOX2, ICOS, TNF-alpha. Methods (11): Serum Proteomics, Spectral Flow Cytometry, RRBS, Phospho-Flow Cytometry, Activation-Induced Marker Assay, T-B Coculture Assay, PRNT, CITE-seq, Immunohistochemistry, Multi-color Immunofluorescence. (FRNT, Single-Cell RNA Sequencing now at 2 sources — resolved.) Will thicken with future ingests; Peripheral Helper T Cell / HOPX / TOX2 are Ansari2025-only and will grow when non-dengue Tph papers are ingested. Peripheral Helper T Cell is the most significant thin page given its centrality to the dengue EF model.

• [2026-05-02 lint] B220 near-orphan — RESOLVED. Added [[B220]] to Double-Negative B Cell Related Pages during 2026-05-08 deep lint.

• “Questions Raised” from Wei2007 — three non-trivial Watch Items:

  • Do DN B cells in dengue share T-bet/FcRL5/CD11c transcriptional identity with atypical B cells in malaria/COVID-19? (Needs comparative phenotyping paper. Sanz2025 provides cross-disease context but no dengue-specific data.)
  • Is the ~3% vs ~5% SHM gap between DN and CD27⁺ memory cells reproducible across disease contexts and VH families? (Wei2007 used VH3 only, n=28.)
  • Can IgD⁻CD27⁻ gating from Wei2007 serve as a retrospective reference gate for re-analysing older dengue flow cytometry datasets? (Needs dengue flow cytometry paper with IgD/CD27 staining.)

• Pre-GC (Bm2ʹ) expansion in dengue unstudied. Anolik2004 identifies circulating IgD⁺CD38^high pre-GC cells (Bm2ʹ) as a distinct SLE-associated population resistant to rituximab. No dengue paper has yet systematically tracked this gate. Does acute dengue activate concurrent GC reactions alongside the dominant EF plasmablast wave? Needs dengue paper with CD38/IgD Bm classification or CD10 staining.

• GC tolerance checkpoint in dengue. Anolik2004 shows that autoreactive VH4.34 memory B cells are expanded in SLE (failure of GC censoring) and normalise after rituximab. Does acute dengue produce a transient GC tolerance failure, generating autoreactive or cross-reactive memory B cells? Needs BCR repertoire data from dengue patients with paired acute/convalescent samples.

• acN cell phenotype in dengue unstudied. Tipton2015 defines activated naive B cells (CD19^hi, IgD⁺, MTG⁺, CD24⁻, CD21⁻, CD23⁻) as major EF ASC precursors in SLE. Does this population expand in acute dengue? MTG is non-standard but CD19^hi + CD21⁻ + CD24⁻ within IgD⁺CD27⁻ cells could be assessed in existing dengue flow cytometry data. Needs dengue paper with naive compartment sub-gating.

• Origin of dengue-specific vs. non-specific plasmablasts. In SLE, naive cells are the dominant ASC precursors during flares (not memory recall). Does a similar naive cell contribution dominate the dengue plasmablast wave, or are pre-existing cross-reactive memory cells the primary precursors? This distinction has implications for whether the dengue plasmablast response is truly EF-naive or EF-memory. Needs dengue connectivity NGS study.

• DN2 expansion kinetics in dengue. Do DN2 cells (CXCR5⁻CD11c⁺CD21⁻) expand during acute dengue? At what kinetics relative to the plasmablast wave? The panel requirements (CXCR5 within IgD⁻CD27⁻ gate) mean this has likely never been measured. Needs dengue flow cytometry paper with CXCR5 staining.

• DN1/DN2 ratio as severity biomarker. Is the DN1/DN2 ratio within IgD⁻CD27⁻ informative of disease severity or EF vs. GC dominance in dengue? Requires CXCR5+CD21+CD11c staining within the DN gate.

• TLR7-driven EF pathway in dengue. Dengue is an ssRNA virus — abundant TLR7 ligands during viraemia. Does the Jenks2018 TLR7/IFN-γ/IL-21 differentiation programme operate in dengue? The antagonistic regulation (TLR7 suppresses GC) predicts EF dominance during active viraemia.

• T-bet/ZEB2 signature in dengue scRNA-seq. Can the T-bet/ZEB2 transcriptional signature be used retrospectively to identify EF-derived B cells in existing dengue single-cell RNA-seq datasets? This is a low-cost computational test of whether DN2-like cells exist in dengue.

• [2026-05-03] Dengue “AtB/ABC” studies need IgD audit. Sanz2025 argues that any study reporting “atypical B cell” expansion without IgD in the panel cannot distinguish DN2 from activated memory or pre-GC contamination. When ingesting dengue papers, check whether IgD was included — flag studies that lack it.

• [2026-05-03] EF/GC endotype concept — testable in dengue? Sanz2025 shows SLE patients segregate into EF-dominant vs. GC-dominant clusters predicting severity and vaccine response quality. Does dengue patient heterogeneity follow a similar endotype pattern? Needs dengue cohort with both DN/plasmablast and GC (CD10⁺/Bm2ʹ/Tfh) markers.

• [2026-05-03] Self-limited EF autoreactivity in dengue. Healthy COVID-19 subjects generate transient naïve-derived autoreactive DN2 cells that resolve within months. Does acute dengue produce a similar transient autoreactive EF response? Does secondary infection perpetuate these clones? Needs dengue acute/convalescent paired BCR data with autoreactivity assays.

• [2026-05-03] DN3 tracking in dengue. DN3 cells (CXCR5⁻CD21⁻CD11c⁻T-bet⁻ within IgD⁻CD27⁻) are pre-plasmablasts expanded in COVID-19 and SLE. Does this population expand in acute dengue alongside DN2? Requires CD11c staining within the DN gate.

• [2026-05-03] ZEB2 as EF predictor in dengue. ZEB2 represses Mef2b (GC TF) → mechanistic basis for EF/GC antagonism. Is ZEB2 elevated in acute dengue B cells? Testable via scRNA-seq re-analysis.

• [2026-05-04] Neutralizing Ab paradox — testable in dengue? Woodruff2020 shows high neutralizing anti-SARS-CoV-2 titers from EF-derived ASCs correlate with poor COVID-19 outcomes. Does an analogous paradox exist in dengue — where high early anti-DENV titers from EF responses correlate with severity or facilitate ADE? This is testable with paired B cell phenotyping + serum neutralization/ADE assays in acute dengue cohorts.

• [2026-05-04] CXCR5/CXCR3 chemokine switch in dengue. The CXCR5↓/CXCR3↑ switch on EF B cells in COVID-19 is a readily measurable marker of EF pathway activation. Can this be detected in acute dengue PBMC samples? Requires CXCR5 and CXCR3 in the flow panel — check whether dengue papers include both.

• [2026-05-04] Woodruff2020 Table 1 as panel design reference. The standardised 24-marker spectral panel with complete B cell population definitions (Table 1) is the reference standard for designing a dengue EF study. Evaluate whether existing dengue flow cytometry studies use compatible gating strategies.

• [2026-05-06] Singh2026 — DN1/DN2/DN3 identity of DENV-specific atypical MBCs unknown. The DENV-specific CD27⁻CD21⁻ cells that accumulate in 2° dengue cannot be subdivided by the Singh2026 panel (lacks CXCR5, CD11c). Are they DN2 (EF effectors) or DN1 (GC memory)? Needs a dengue study with CXCR5/CD11c within the DN gate.

• [2026-05-06] Singh2026 — IgM+ MBC recall mechanism in 2° dengue. IgM+ MBCs are the only subset significantly elevated during acute secondary dengue. Are these EF-derived (low SHM, broad cross-reactivity per Tipton2015 model) or GC-derived? BCR sequencing of sorted IgM+ DENV-specific cells would resolve this.

• [2026-05-06] Singh2026 — delayed MBC peaks and 12–18M uptick. The >3-month delayed peaks in 2° dengue and the 12–18M uptick in 3/4 secondary cases could reflect prolonged GC reactions, tissue redistribution, or subclinical re-exposure in endemic settings. Needs longitudinal study with GC markers (CD10, Tfh) or serology to rule out re-exposure.

• [2026-05-06] Singh2026 — naïve-like IgD+/IgM+ DENV-specific cells. These persist to 18M in both 1° and 2° infection. Their identity is ambiguous: true antigen-experienced cells with SHM, germline-encoded polyreactive B cells, or activated naive precursors. BCR sequencing of this population would resolve whether they carry SHM or are germline.

• [2026-05-06] Singh2026 — qualitative MBC reprogramming and functional outcomes. Does the shift toward IgG+/atypical/class-switched MBC subsets in 2° dengue predict neutralizing antibody breadth, ADE-relevant cross-reactivity, or clinical severity upon subsequent infection?

• [2026-05-07] AP-1/EGR chromatin remodelling in dengue? Scharer2019 identifies AP-1/EGR motif amplification as the SLE-specific disease layer on DN2 chromatin (distinct from the shared T-BET programme). Does acute dengue (an ssRNA virus with TLR7 activation and IFN-γ) produce transient AP-1/EGR remodelling? If so, does it resolve post-defervescence or persist? Testable by ATAC-seq on sorted B cells from acute dengue patients.

• [2026-05-07] ATF3 as EF pathway marker in dengue. ATF3 is induced by both TLR stimulation and cellular stress — both prominent in dengue. Intracellular ATF3 flow cytometry (validated in Scharer2019) could be added to dengue panels as a practical readout of EF pathway activation. Needs testing in dengue PBMC samples.

• [2026-05-07] DN2 apoptosis resistance in dengue. Scharer2019 shows DN2 cells uniquely lack G2/M checkpoint and apoptosis pathway enrichment among SLE B cell subsets. Does a similar mechanism explain atypical/DN MBC persistence in secondary dengue (Singh2026)?

• [2026-05-07] Naive B cell epigenetic priming in dengue-endemic settings. SLE resting naive cells carry 6,664 DMLs and NR4A1/NR4A3 upregulation (BCR+TLR engagement). Does chronic dengue exposure in endemic settings produce analogous epigenetic priming of naive B cells, lowering the EF activation threshold?

• [2026-05-07] 111-CpG biomarker signature — cross-disease applicability? Scharer2019 identifies 111 CpGs that discriminate SLE from healthy B cells across all subsets. Could these serve as biomarkers for EF pathway activation in other diseases with EF dominance, including severe dengue?

• [2026-05-08 lint] External Citation Audit — 54 external papers cited inline across ~25 pages. See analyses/External Citation Audit.md. Curator reviewing DOIs and accuracy. Once verified: (1) rewrite inline citations to attribute solely to ingested sources, and/or (2) ingest high-priority external papers. 3 bare citations (Pattern A) are highest priority to resolve.

• [2026-05-08 deep lint] Content-after-Related-Pages displacement — FIXED. 19 pages had Key Points bullets appended after ## Related Pages. All moved into correct section. Root cause: ingest workflow was appending at end-of-file instead of inserting into Key Points section.

• [2026-05-08 deep lint] Template compliance — FIXED. Added ## Contradictions & Debates to 43 pages that were missing it. All pages now have complete section structure.

• [2026-05-08 deep lint] Frontmatter/index count mismatches — FIXED. FACS Sorting, In Vitro B Cell Stimulation, ELISpot frontmatter corrected. Index header counts corrected (Entities 41, Methods 17, Total 79).

• [2026-05-08 deep lint] Source page cross-reference cleanup — FIXED. Memory B Cell removed from Entities Mentioned in 5 sources; Singh2026 FACS Sorting link removed; Wrammert2012/GarciaBates2013 concept links added; Research Plan broken wikilinks fixed; Notable Findings heading repositioned; B220/CD10 orphan links added.

• [2026-05-08] ⚠ Ansari2025 ingest disrupted by API errors — DEEP LINT COMPLETED. Checked all 22 entity, 5 concept, and 8 method pages linked from Ansari2025. Found and fixed 4 issues: (1) T-bet entity page was never visited — added source line, Key Points bullet, frontmatter update; (2) Germinal Center missing Key Points bullet about CXCL13/concurrent GC — added; (3) Memory B Cell missing Key Points bullet about Tph→memory B cell preference — added; (4) Conventional Flow Cytometry missing panel details + wrong frontmatter count — added panel bullet, fixed sources 8→9. BLIMP-1 was NOT listed in Ansari2025’s Entities Mentioned (not a propagation failure). All other pages verified clean.

• TLR7-driven EF pathway in dengue. Partially resolved by Ansari2025: IL-21-dependent Tph mechanism provides the T cell help arm. TLR7 (B cell-intrinsic) not directly measured but the IL-21 axis maps onto the Jenks2018 pathway.

• DN2 expansion kinetics in dengue. Partially resolved: Ansari2025 shows CD21⁻CD11c⁺ (DN2-phenotype) B cells expanded in acute dengue. Kinetics relative to PB wave not detailed; formal DN2 confirmation (T-bet, CXCR5) still needed.

• Neutralizing Ab paradox — testable in dengue? RESOLVED: Ansari2025 replicates the paradox — anti-NS1/anti-prM IgG elevated in severe dengue but FRNT₅₀ not different by severity.

• [2026-05-08] Ansari2025 — are CD21⁻CD11c⁺ B cells in dengue truly DN2? T-bet, CXCR5, and FCRL5 staining were not performed within the DN gate. Formal DN2 confirmation requires intracellular T-bet + CXCR5 surface staining on IgD⁻CD27⁻ cells from acute dengue samples.

• [2026-05-08] Ansari2025 — Tph→ADE link. Does the Tph→IL-21→memory B cell→PB axis produce ADE-competent cross-reactive IgG in secondary dengue? This is the key translational question. Needs paired Tph frequency + ADE assay data.

• [2026-05-08] Ansari2025 — cytotoxic Tph function unknown. The GZMB⁺HOPX⁺ Tph subcluster has no assigned function. Does it kill B cells, infected cells, or regulate the immune response? Only 13 shared TCR clonotypes with helper Tph.

• [2026-05-08] Ansari2025 — IL-21 as therapeutic target. Anti-IL-21 reduces PB output ~60%. Does this reduce severity-associated non-neutralizing IgG while preserving neutralizing titers? Testable in vitro with dengue-specific readouts.

• [2026-05-08] Ansari2025 — EF+GC concurrent activity. CXCL13 elevated alongside Tph dominance suggests both pathways active simultaneously. Is the EF/GC ratio a severity biomarker? Needs combined Tph + Tfh + CXCL13 + DN2:DN1 assessment.

• [2026-05-08] Research Plan update needed. RESOLVED: Revision 2 (2026-05-08) incorporates Ansari2025 throughout. Rationale reframed, H4 added, Tph quantification added as follow-up study 7. Tph staining (CXCR5/PD-1) kept as follow-up rather than primary panel (requires second tube). IL-21 measurement and FRNT remain out of scope for the flow-only B cell study but are noted as complementary readouts.

• [2026-05-08] GarciaBates2013 — functional quality of plasmablast antibodies — RESOLVED. Priyamvada2016 provides the monoclonal characterisation: 70% E-specific, all cross-reactive, 45/53 ADE-competent, OAS in 2/4 patients. The PB wave produces IgG that is predominantly cross-reactive, ADE-competent, and in some patients serotype-biased by OAS.

• [2026-05-08] GarciaBates2013 — age/interval effect on serotype cross-reactivity. Brazilian adults (>20y interval) show infecting-serotype-dominant reactivity; Nicaraguan children (shorter interval) show previous-serotype-dominant (original antigenic sin). Is this interval-dependent, age-dependent, or serotype-specific? Needs a cohort study with controlled interval comparisons.

• [2026-05-08] GarciaBates2013 — memory vs. naive origin of dengue plasmablasts. Anamnestic kinetics (days 4–7 peak, secondary >> primary) favor memory recall, consistent with Ansari2025 Tph→memory B cell mechanism. But the naive B cell contraction in 2° DFC could also reflect naive recruitment. Needs connectivity NGS or clonal tracking.

• [2026-05-08] GarciaBates2013 — B cell apoptosis mechanism and pathological significance. 60% caspase-3⁺ B cells in 2° DFC with Ki-67/caspase-3 and CD95/caspase-3 correlations. Is this Fas-mediated AICD? Does it limit the plasmablast wave (homeostatic) or contribute to pathology? Connects to DN2 apoptosis resistance (Scharer2019) — are DN2-phenotype cells selectively spared?

• [2026-05-08] GarciaBates2013 — paracrine IL-21 missed by serum assay? Authors found no correlation between serum IL-21 and plasmablast frequency. Ansari2025 (12 years later) identified Tph-derived paracrine IL-21 as the driver. This suggests bulk serum cytokine assays miss the relevant compartmentalized signal — FluoroSpot or intracellular staining needed.

• [2026-05-08] Wrammert2012 — Fc glycosylation of dengue PB-derived IgG unstudied. The paper raises IgG Fc glycosylation patterns as a potential pro-inflammatory mechanism but provides no data. Given that afucosylated IgG enhances FcγRIIIa binding and ADCC, and that the massive dengue PB wave produces predominantly non-neutralizing IgG, the glycosylation state of these antibodies could directly influence whether they are protective or pathological. Needs dengue study with Fc glycoproteomics on PB-derived mAbs.

• [2026-05-08] Parameswaran2013 — antigen specificity of convergent CDR3s unknown. The convergent CDR3s (ARLDYYYYYGMDL etc.) are shared across individuals and serotype-independent, but their target antigen is unknown. Do they bind DENV E protein, NS1, prM, or other targets? If cross-serotype-conserved epitopes, are these antibodies neutralizing or ADE-enhancing? This is directly testable by expressing recombinant antibodies with these CDR3s and testing binding/neutralization.

• [2026-05-08] Parameswaran2013 — convergent CDR3s not cell-type resolved. These CDR3s were identified from unsorted PBMC gDNA. Whether they are carried by plasmablasts, memory B cells, or both is unknown. Sorting acute-phase B cells (plasmablasts vs. memory vs. naive) before BCR sequencing would determine which compartment harbours the convergent clones and connect to the Tph→memory B cell→PB pathway (Ansari2025).

• [2026-05-08] Parameswaran2013 — intermediate SHM does not resolve EF vs. GC origin. The 4.4–6.9% V mutation in convergent CDR3-bearing cells falls between the EF (<3%) and GC (~7.3%) benchmarks from SLE (Tipton2015). Three models remain: (a) GC-matured memory cells recalled via EF pathway; (b) mixture of EF and GC populations; (c) more extensive EF maturation in dengue than SLE. Sorted plasmablast BCR sequencing would resolve this — the key remaining gap.

• [2026-05-08] Wrammert2012 — primary vs. secondary PB kinetics underpowered. Only 4/46 were primary infections; responses appeared similar but too few for comparison. This remains a gap — no study in the wiki has adequately powered primary vs. secondary PB kinetics data. The distinction matters because primary responses should have more IgM and potentially different kinetics if naive-derived (EF) vs. memory-derived.

• Dengue plasmablast kinetics benchmark — RESOLVED. Wrammert2012 provides the kinetics: barely detectable at days 2–3, rapid increase to peak at day 6–7, return to baseline by 1 month post-discharge. This matches the GarciaBates2013 days 4–7 peak window. The decline post-defervescence kinetics (days 7–14–21) remain incompletely characterised — return visits were at ≥1 month, not at intermediate convalescent timepoints.

• [2026-05-09] Appanna2016 — VH4-34/VH1-69 autoreactivity in dengue PBs. PBs were enriched for VH4-34 and VH1-69 (autoantigen-associated V genes). Is this transient EF autoreactivity (as in COVID-19, Sanz2025) or persistent? Does it contribute to dengue-associated autoimmune phenomena? Needs paired acute/convalescent BCR data.

• [2026-05-09] Appanna2016 — IgM-only shared PB/MBC clones — artefact or biology? The rare CDR3s shared between PBs and MBCs were exclusively IgM. These could be genuine cross-compartment lineage members (evidence for IgM memory → PB pathway) or low-affinity polyreactive IgM artefacts from the live-virus sorting approach. Single-cell paired VH/VL with antigen validation would resolve this.

• [2026-05-09] Appanna2016 — what activates prM-specific and complex-epitope MBCs? PBs are 85% E-specific, but MBCs are primarily complex-epitope and prM-specific. If the Tph→memory B cell→PB axis selectively activates E-specific memory, what drives the prM/complex-epitope MBC response? A separate activation pathway (GC? Bystander?) must be invoked.

• [2026-05-09] Appanna2016 — CD27⁺ MBC gate misses DN memory. The study gated MBCs as CD19⁺CD20⁺CD27⁺, excluding the entire IgD⁻CD27⁻ (DN) compartment. Given that DN/atypical MBCs accumulate in secondary dengue (Singh2026) and are expanded acutely (Ansari2025), the “DENV-binding MBC” population in this study is incomplete. A study using IgD/CD27 gating with antigen probes would capture the full DENV-specific memory repertoire.

• [2026-05-09] GodoyLozano2016 — antigen specificity of low-SHM IgG unknown. Are the germline-coded, cross-reactive IgG antibodies DENV-specific? The polyclonal CDRH3 diversity within biased IGHV segments suggests germline-encoded recognition, but no antigen-specific sorting was performed. Needs paired BCR sequencing + antigen specificity testing.

• [2026-05-09] GodoyLozano2016 — convergent CDRH3 epitope targets unknown. The convergent CDRH3s (ARQFGNWFDS, ARQWGNWFDL) shared across 52% of patients have unknown antigen targets — same gap as Parameswaran2013 convergent CDR3s. Recombinant antibody expression + binding/neutralization assays would resolve this.

• [2026-05-09] GodoyLozano2016 — why is SHM lower in secondary than primary? Two competing models: (1) stronger innate-like EF response in secondary infection (authors’ proposal); (2) original antigenic sin rapidly activates low-SHM cross-reactive memory clones, outcompeting high-SHM serotype-specific clones. Distinguishing these requires sorted PB BCR sequencing with lineage tracing.

• [2026-05-09] GodoyLozano2016 — TLR7-mediated T-independent CSR not experimentally demonstrated. The proposed mechanism (endosomal DENV ssRNA → TLR7 → AID → CSR without GC SHM) is biologically plausible but entirely inferred. No experiment in this study (or any dengue study in the wiki) directly tests TLR7 activation in dengue B cells. Needs in vitro dengue B cell stimulation with TLR7 agonists/antagonists.

• [2026-05-09] GodoyLozano2016 — IGHV1-2/1-69 and ADE. Is the association between IGHV1-2 low SHM and disease severity (DWS+) causal? Do IGHV1-2- and IGHV1-69-using antibodies preferentially mediate ADE? Needs functional characterization of V-gene-stratified dengue antibodies.

• [2026-05-10 lint] Evidence weight annotations missing in early-ingested pages. ~15 Key Points bullets across Wei2007 and early Tipton2015 ingests lack study type and sample size annotations (e.g., “n=29 cross-sectional”). Pages affected: CD24, IgD, CD10, IgM, IgA, B220, Double-Negative B Cell, CD38, CD27. Concentrated in the first 2–3 source ingests; later ingests are well-annotated. Low priority but should be addressed during a quiet session.

• [2026-05-10 lint] CD24 lists Scharer2019 in Sources with no Key Points content. Source was added during Scharer2019 ingest (CD24 is mentioned) but no bullet was written. Should add a brief note about CD24⁻ epigenetic status in DN2/aNAV from Scharer2019 data.

• [2026-05-10 lint] Wrammert2012 ELISA listed as plain text — no method page. ELISA is listed under Methods Used without a wikilink. Either create an ELISA.md method page or convert to plain text note. Low priority — ELISA is a standard technique, not a distinguishing method for this wiki’s scope.

• [2026-05-10 lint] SHM contradiction surfaced: GodoyLozano2016 vs. Appanna2016. Added to Somatic Hypermutation Contradictions & Debates with reconciliation (unsorted IgG pool vs. FACS-sorted DENV-specific populations). No further action unless new SHM data from another dengue paper changes the picture.

• [2026-05-10] Priyamvada2016 — OAS variability across patients. OAS (preferential DENV1 neutralisation during DENV2 infection) was present in only 2/4 patients. Is OAS universal in secondary dengue or dependent on serotype combination, interval between infections, or individual clonal history? Needs larger cohort with known prior serotype exposure.

• [2026-05-10] Priyamvada2016 — GC vs. EF origin of high-SHM PBs. The high SHM (mean 18.1 VH) is consistent with GC-experienced memory recall, but does not exclude EF-matured memory. Do the memory B cells that give rise to these high-SHM PBs originate from prior GC reactions, or can EF responses accumulate comparable SHM over multiple exposures? Lineage tracing across primary/secondary paired samples would resolve this.

• [2026-05-10] Priyamvada2016 — OAS mAbs as ADE mediators. The DENV1-biased mAbs bind DENV2 weakly and fail to neutralise it — the precise profile for ADE. Are these OAS mAbs specifically the subset that mediates ADE in secondary DENV2 infection? Needs ADE assays stratified by OAS vs. non-OAS mAbs.

• [2026-05-10] Priyamvada2016 — in vivo relevance of universal ADE. 45/53 mAbs enhanced at 1 µg/ml in vitro (U937). Is this clinically meaningful? The ADE assay used a single concentration; concentration-dependent neutralisation-to-enhancement transitions were not characterised. Needs dose-response ADE curves.

• [2026-05-09] SHM distribution of dengue acute-phase plasmablasts — SUBSTANTIALLY RESOLVED. Priyamvada2016 provides sorted PB BCR sequencing: mean 18.1 VH mutations in secondary DHF PBs, supporting memory origin. Combined with GodoyLozano2016 low SHM in unsorted IgG, the dual-pathway model (memory recall + de novo EF) is now the best framework. Remaining gap: primary dengue sorted PB SHM data.

• Web deployment — DEPLOYED. Site live at efb-dengue-wiki.pages.dev via Cloudflare Pages (GitHub repo OsandaC/efb-dengue-wiki, branch main). Quartz setup in efbwebshare/ folder (sibling of efb-dengue-wiki/). Update Web workflow added to CLAUDE.md. Run sync-and-build.ps1 from efbwebshare/ to deploy. [2026-05-14] Sync script updated to also copy Claude-council/ → content/council/ for web publication of council reports. [2026-06-15] Ran update web (118 pages, Lamprinou2026 + B Cell Panel Variant 1 + Thesis Objectives + Compensation/FMO + DN2 Panel SOP pages now live). Push initially rejected (non-fast-forward) — GitHub had a merged dependabot PR (a23275e, production deps) ahead of local main. Resolved via git fetch + git rebase origin/main + push (no conflicts). [2026-06-29] Ran update web — 121 content files → 574 emitted; pushed cleanly (efbwebshare main dded26b, no conflicts). This deploy caught up both the 2026-06-27 Switched Memory page and the Narvaez2011 ingest. Watch: if sync-and-build.ps1’s push fails with “fetch first,” the cause is likely a merged dependabot/GitHub-side PR — fetch and rebase efbwebshare’s main before retrying.

• [2026-05-14 council, fixed 2026-05-23] Ansari2025 — “Tph” identity vs. canonical Tph (Rao2017). ✅ Th1 signature mismatch caveat added to EF Response page Dengue Context (first Tph mention). These CXCR5⁻PD-1⁺ cells have Th1 signatures (CXCR3⁺, T-bet, IFN-γ), not canonical Tph (MAF⁺, CXCL13⁺). Remains an open research question — a paper comparing dengue Tph transcriptomes to RA/SLE Tph would resolve this.

• [2026-05-14 council, fixed 2026-05-23] Ansari2025 — memory vs. naive precursor tension. ✅ Coculture caveat added to EF Response and Memory B Cell pages. The memory B cell preference may partly reflect the experimental system (seropositive donor memory T cells; acute-phase T cells died in culture). Remains an open research question.

• [2026-05-14 council, fixed 2026-05-23] Ansari2025 — day-of-sampling confounder for severity. ✅ Caveat added to EF Response page Tph severity bullet. Severe patients sampled later (median 8 vs 5 days). Remains an open research question — multivariable adjustment needed in future studies.

• [2026-05-14 council, fixed 2026-05-23] Ansari2025 — CXCL13 is not GC-specific. ✅ GC page CXCL13 bullet rewritten with full hedging (was HIGH contradiction with EF page). “Suggests GC reactions are simultaneously active” → “cannot be used as evidence of concurrent GC activity without independent GC markers.”

• [2026-05-14 council] Ansari2025 — IL-21 blocking lacks parallel Tfh arm. The IL-21R-Fc blocking (Figure 6J) was done only with CXCR5⁻PD-1⁺ cells. Whether Tfh-driven PB generation is also IL-21-dependent (or uses a different cytokine) remains untested.

• [2026-05-14 deep lint] Dual-pathway model framing — “resolves” → “working hypothesis”. ✅ Fixed on EF Response page 2026-05-18 (council-directed). Multiple lint agents flagged overclaiming. The dual-pathway model (memory-derived high-SHM PBs + de novo EF low-SHM) is a working hypothesis, not a resolution. Affects: Extrafollicular Response, Somatic Hypermutation, state.md Current Focus. Curator to decide scope of softening.

• [2026-05-14 deep lint, partially resolved 2026-05-24] “DN2” vs. “DN2-phenotype” systematic rewrite. ✅ Research Plan (Rev 4) fully updated: all DN subdivision references now use “DN2-phenotype”/“DN1-like”/“DN3-like” terminology. DN2 Gating Strategy analysis already uses correct terminology. Remaining: Entity/concept pages (Double-Negative B Cell, DN2 B Cell, Extrafollicular Response, etc.) still use “DN2” where only Ansari2025 is the dengue source — systematic audit of wiki body pages not yet done.

• [2026-05-14 deep lint] IFN-gamma entity page missing. Referenced across 5+ pages (T-bet signalling, CXCR3 induction, Tph cytokine profile) but lacks its own entity page. Low priority — could be created during next ingest that mentions IFN-gamma directly.

• [2026-05-18] William2002 — does EF SHM rate translate to human responses? The murine EF SHM rate (~0.3 mut/Vκ/generation) is the only quantitative estimate of mutation rate at EF sites. Whether human DN2-derived SHM accumulates at a comparable rate is unknown — human data (Jenks2018) compares total mutation levels between populations, not per-generation rates. This question is relevant to modelling how quickly dengue EF responses could accumulate SHM during acute infection.

• [2026-05-18] William2002 — TLR9→TLR7 mechanistic analogy for dengue. The paper proposes TLR9 co-stimulation (via chromatin immune complexes) as the driver of sustained EF B cell activation. The direct analogy is TLR7 sensing of DENV ssRNA during viraemia. No study has tested whether TLR7 stimulation produces sustained EF B cell proliferation + SHM in the context of dengue. In vitro dengue B cell stimulation with R848 (TLR7 agonist) would test this.

• [2026-05-18] William2002 — EF tolerance escape and dengue autoreactivity. The paper proposes that EF mutation escapes GC tolerance checkpoints because FDCs and Fas-mediated apoptosis are absent at EF sites. If this mechanism operates in dengue, it could explain the VH4-34/VH1-69 autoreactive enrichment in dengue plasmablasts (Appanna2016) and the transient autoreactivity model (Sanz2025). Testing requires paired BCR sequencing + autoreactivity assays on sorted acute dengue B cells.

• [2026-05-19] Bhattacharya2016 — tissue-resident PBs in secondary dengue? The commentary raises the possibility that additional PB subsets form during secondary dengue but are retained in secondary lymphoid organs rather than circulating. If so, circulating PB analysis (Wrammert2012, GarciaBates2013) would undercount the total PB response, and the tissue-retained subset might have different specificities (e.g., prM/NS rather than E). No dengue study in the wiki addresses this — would require paired blood + tissue sampling.

• [2026-05-14 deep lint] Evidence weight annotations — early-ingested pages. ~33% of Key Points bullets across Wei2007/Tipton2015-era entity pages lack study type and sample size annotations. Concentrated in CD24, IgD, CD10, IgM, IgA, B220, Double-Negative B Cell, CD38, CD27. Low priority maintenance task.

• [2026-05-14 deep lint, resolved 2026-05-23] OAS/ADE pages — thickened. OAS: 2→7 sources (added GodoyLozano2016, Parameswaran2013, Appanna2016, Bhattacharya2016, Ansari2025, GarciaBates2013). ADE: 2→5 sources (added GarciaBates2013, Ansari2025, Woodruff2020). GarciaBates2013 added as OAS counter-evidence (infecting-serotype dominance in Brazilian cohort).

• [2026-05-14 deep lint] Missing cross-folder links (method↔concept). Method pages rarely link to concept pages and vice versa (e.g., BCR Sequencing ↛ Somatic Hypermutation). Low priority structural improvement.

• [2026-05-15 council] Woodruff2020 — CXCR3 entity page should note pre-PB vs. mature ASC distinction. Woodruff2020 documents CXCR5↓/CXCR3↑ on pre-PB EF populations (aN, DN2), not on mature ASCs. The CXCR3 entity page currently does not make this distinction — relevant for interpreting Ansari2025 dengue CXCR3 data.

• [2026-05-15 council] Woodruff2020 — naive vs. memory EF precursors cross-disease gap. The naive-derived, germline-dominant EF response in COVID-19 (primary infection) contrasts with memory-dominated, high-SHM response in secondary dengue (Priyamvada2016). Wiki should explicitly flag this distinction: Woodruff2020 benchmarks what a primary EF ASC response looks like. Whether primary dengue produces a similar germline-dominant pattern is untested.

• [2026-05-15 council, updated 2026-05-22] Sanz2025 — memory DN2 cells in dengue? Sanz2025/Faliti2024 show antigen-specific DN2 cells persist >1 year post-SARS-CoV-2 vaccination (>50% of spike/RBD⁺ cells), establishing durable CD27⁻ memory with a DN2 phenotype. Singh2026’s DENV-specific CD27⁻CD21⁻ MBCs persisting to 18 months in secondary dengue could represent memory DN2 rather than effector DN2. Sutton2021 update: The MBC1 cluster provides transcriptomic evidence for a quiescent “memory DN2” population within the alternative lineage — partially confirming the Sanz2025/Faliti2024 prediction. Whether dengue generates MBC1-equivalent cells remains untested.

• [2026-05-15 council] Sanz2025 — context-dependence principle for Ansari2025 secondary cohort. Sanz2025 insists that CD11c⁺T-bet⁺CD21lo cell identity depends on context (primary vs. recall). Ansari2025’s dengue cohort is predominantly secondary infections — Sanz2025 would predict these CD21⁻CD11c⁺ cells could be memory-derived DN2-like cells rather than canonical naïve-derived DN2 (as in SLE primary flares). This weakens the direct SLE analogy and reinforces the existing watch item about memory vs. naive precursor tension.

• [2026-05-15 council] Jenks2018 — BCR-independent DN2→PC as bystander PB mechanism. ✅ Added to EF Response page 2026-05-18. Figure 7F shows DN2 cells generate plasmablasts without BCR cross-linking. This provides a mechanistic explanation for the non-DENV-specific (bystander) fraction of the dengue PB wave — not currently documented in any wiki page. When wiki pages are next updated for Jenks2018, this should be added to the Extrafollicular Response concept page and the Plasmablast entity page.

• [2026-05-15 council] Jenks2018 — IgG3 enrichment on DN2 cells. DN2-derived PBs are IgG3-enriched in SLE. IgG3 is the most complement-activating subclass. Whether this applies in dengue is unknown and potentially relevant to pathogenesis (complement activation, vascular leak). Track as a novel angle for dengue EF studies.

• [2026-05-15 council] Sanz2025 — DN classification is Sanz lab framework, not independently validated. The council rated the central thesis (abandon AtB for DN1/DN2/DN3) as WEAK evidence — conceptual argument, not experimental validation. Wiki uses DN nomenclature (per Decision 2026-05-02) and this remains the most precise available, but should be acknowledged as the Sanz lab’s framework pending cross-lab validation. No wiki changes needed unless alternative classifications emerge.

• [2026-05-17 council] Extrafollicular Response page — 5 displaced bullets in Contradictions & Debates. ✅ Fixed 2026-05-18. Singh2026 (×2), Appanna2016 (×1), GarciaBates2013 (×2) positive-finding bullets are displaced into ## Contradictions & Debates from end-of-file insertion during past ingests. All five belong in ## Dengue Context. This is the same insertion-order defect fixed in the 2026-05-08 deep lint, re-occurring on the highest-traffic concept page. Curator to direct fix.

• [2026-05-17 council] Extrafollicular Response page — “resolves the SHM paradox” FATAL FLAW. ✅ Fixed 2026-05-18. The Priyamvada2016 Dengue Context bullet ends with “This resolves the SHM paradox: the EF pathway is real…” — the dual-pathway model is a working hypothesis, not a resolution (n=4 + n=19, non-overlapping populations, neither study designed to test the model). Replace “resolves” with “offers a plausible reconciliation of” and add explicit hypothesis caveat.

• [2026-05-17 council] Extrafollicular Response page — dengue EF mechanism divergence unflagged. ✅ Fixed 2026-05-18. The page describes the EF pathway as TLR7 + IFN-γ + IL-21 (SLE/Jenks2018 model) throughout Key Points, then describes dengue as Tph-dependent (IL-21-mediated) in the Dengue Context — but never flags that these are mechanistically different. A dengue reader will incorrectly infer TLR7-autonomous B cell activation in dengue. Needs one sentence in Dengue Context noting the Tph-dependent vs. TLR7-autonomous distinction.

• [2026-05-17 council] Extrafollicular Response page — BCR-independent DN2→PC absent. ✅ Fixed 2026-05-18. The mechanistically critical finding (Jenks2018 Figure 7F: DN2 cells generate plasmablasts without BCR stimulation) is documented on the DN2 B Cell and Plasmablast pages but absent from the EF Response page. This is the hinge for explaining bystander plasmablast generation in dengue.

• [2026-05-17 council] Extrafollicular Response page — memory DN2 cells absent. ✅ Fixed 2026-05-18. Sanz2025/Faliti2024 establish durable antigen-specific DN2 cells persisting >1 year (memory DN2). The page frames EF output as exclusively short-lived throughout. The effector DN2 vs. memory DN2 distinction should be added.

• [2026-05-17 council] Extrafollicular Response page — CXCR3 missing from Related Pages. ✅ Fixed 2026-05-18. CXCR3 appears inline in the Woodruff2020 Key Points bullet but is absent from Related Pages. Add CXCR3 to Related Pages.

• [2026-05-17 council] GodoyLozano2016 — “385,206 lineages” unverifiable in source page. ✅ Verified against PDF 2026-05-18; added to source page. This specific dataset-size claim appears in the EF Response concept page but not in the GodoyLozano2016 source page. Needs verification against original PDF before it can be trusted. If incorrect, correct in both pages.

• [2026-05-17 council] Jenks2018 — per-cell IgG claim unverifiable in source page. ✅ Removed 2026-05-18; replaced with verified ELISPOT data from source page. The EF Response page states DN2 cultures “produce IgG at higher per-cell levels than DN1 or SWM.” The Jenks2018 source page does not contain this; it records that surface IgG is 50% lower on DN2 than SWM/DN1 (surface ≠ secreted, but the claim is ambiguous). Needs verification against original PDF or removal.

• [2026-05-20, updated 2026-05-22] DN2 Gating Strategy — CD21⁻/CXCR5⁻ concordance in dengue unknown. Jenks2018 shows <5–10% discordance in SLE; whether this holds in acute dengue determines whether “DN2-phenotype” approximates true DN2. No dengue data exists. Sutton2021 update: CITE-seq shows CD21⁻CD27⁻ captures only 44.7% of transcriptomic alternative lineage cells — CD11c is the better discriminator. This raises the stakes: CD21-based surrogates may substantially undercount the population of interest.

• [2026-05-20] DN2 Gating Strategy — CD27 shedding magnitude in acute dengue unquantified. ADAM17-mediated CD27 cleavage in high-TNF/IL-6 dengue could inflate apparent DN frequencies. Measuring sCD27 in matched serum samples would quantify contamination risk. No dengue study in the wiki addresses this.

• [2026-05-20] DN2 Gating Strategy — second-tube validation panel needed. Adding CXCR5, intracellular T-bet, IgM, IgG in a subset of samples would calibrate the primary panel’s DN2 purity and resolve isotype distribution within DN2-phenotype cells.

• [2026-05-22] Sutton2021 — does the alternative lineage framework apply to dengue acute infection? The alternative lineage was characterised in malaria and healthy donors. Dengue acute infection features more explosive plasmablast expansion than malaria — does this shift the balance from alternative memory (MBC1) toward PC differentiation, as Sutton’s Discussion proposes for SLE? No dengue study has performed scRNA-seq on B cells with sufficient depth to test alternative vs. classical lineage representation.

• [2026-05-22] Sutton2021 — how much have dengue studies underestimated alternative lineage cells? CD21⁻CD27⁻ captures only ~45% of transcriptomic atBCs. All dengue studies using this gate (Ansari2025, Singh2026, GarciaBates2013) have underestimated the population. Would CD11c-based gating reveal larger DN2-phenotype expansions in dengue?

• [2026-05-22] Sutton2021 — context-dependent PC fate applicable to severe dengue? Sutton reconciles the atBC-as-pre-PB model (Jenks2018/SLE) with atBC-as-memory model (healthy/infection) via context dependence: chronic TLR7 stimulation may drive PC fate. Severe secondary dengue features high inflammatory cytokines, TLR7 ligands, and immune dysregulation overlapping with SLE. Is severe dengue the context where atBCs are pushed toward PC differentiation?

• [2026-05-22] Sutton2021 — IgG3 enrichment and complement pathology in dengue. Alternative lineage cells are enriched for IgG3 in malaria-exposed donors. IgG3 is the most complement-activating subclass. If this holds in dengue, alternative lineage-derived IgG3 could contribute to complement-mediated vascular leak in severe disease. Relates to Jenks2018 IgG3 enrichment watch item.

• [2026-05-22] Sutton2021 — functional antibody output of MBC1 (alternative memory) upon rechallenge. MBC1 cells are quiescent alternative lineage memory. What happens when they are recalled? Are they a source of cross-reactive or autoreactive antibodies in secondary infection? No functional data exists for this population.

• [2026-05-22] Kaneko2020 — does dengue TNF-α disrupt GC TFH differentiation? Kaneko2020 shows TNF-α accumulation blocks Bcl-6⁺ TFH differentiation in COVID-19. Severe dengue features cytokine dysregulation with elevated TNF-α. If the same mechanism operates, it would mechanistically explain GodoyLozano2016’s low-SHM IgG findings (no GCs → no GC-level SHM). Needs dengue tissue or paired TNF-α + GC marker data.

• [2026-05-22] Kaneko2020 — Bcl-6⁺ TFH in dengue lymph nodes? Kaneko2020 shows complete absence of Bcl-6⁺ GC-TFH in COVID-19 tissue despite preserved CXCR5⁺ pre-GC TFH. Detecting Bcl-6⁺ TFH in dengue-draining LNs would distinguish dengue (potentially concurrent EF+GC) from COVID-19 (GC ablation). No dengue tissue immunofluorescence study exists in the wiki.

• [2026-05-22] Kaneko2020 — AID⁺/Bcl-6⁻ B cell profile in dengue? The preservation of AID⁺ B cells despite loss of Bcl-6⁺ GC B cells in COVID-19 tissue is the strongest evidence that AID operates in EF contexts. Does dengue show a similar dissociation? Intracellular AID staining on dengue B cells (acute phase) would test whether AID expression occurs independently of GC markers.

• [2026-05-22] Kaneko2020 — TNF-α blockade as GC rescue in severe viral infections? The paper proposes that anti-TNF biologics could rescue GC formation. Murine malaria data (Ryg-Cornejo 2016) support this. If dengue GC suppression is TNF-α-mediated, anti-TNF could restore GC SHM and improve antibody quality — but risks enhancing viraemia. Translational question requiring animal model testing.