🧬 Aging Wiki

R11

18 min read

log/R11.md — Round 11 entries

Sub-file of log — see parent for index.

[2026-05-05] ingest | taurine compound page

  • added: molecules/compounds/taurine.md
  • entity type: compound (type: compound)
  • canonical IDs confirmed: PubChem CID 1123, InChIKey XOAAWQZATWQOTB-UHFFFAOYSA-N, CAS 107-35-7, ChEMBL CHEMBL239243, DrugBank DB01956
  • primary DOIs cited:
    • 10.1126/science.abn9257 (Singh 2023 Science — gold OA, archive status: pending download; abstract + Crossref metadata only; NOT PDF-verified)
    • 10.1126/science.adl2116 (Fernandez 2025 Science — closed-access; abstract-only verification; no-fulltext-access)
    • 10.3389/fphys.2021.700352 (Chen 2021 Front Physiol — gold OA, archive status: pending download; abstract-only)
    • 10.1139/apnm-2012-0229 (da Silva 2014 APNM — closed-access; abstract-only; no-fulltext-access)
  • DOI corrections made at seeding:
    • Chen 2021: brief-supplied DOI 10.3389/fphys.2021.756338 corrected to 10.3389/fphys.2021.700352 (confirmed via PubMed PMID 34497536)
    • da Silva 2014: brief-supplied n=9 corrected to n=21 (placebo n=10, taurine n=11) per PubMed abstract PMID 24383513
  • implicit study stubs created (4): singh-2023-taurine-deficiency-aging, fernandez-2025-taurine-aging-biomarker, chen-2021-taurine-exercise-dose-response, dasilva-2014-taurine-eccentric-exercise
  • gaps surfaced: contradictory-evidence (Singh 2023 vs Fernandez 2025 on taurine-decline-with-aging premise); needs-human-replication (no human aging RCT); needs-replication (mouse lifespan extension single-lab); long-term-unknown; dose-response-unclear; no-fulltext-access (Fernandez 2025, da Silva 2014)
  • Singh 2023 PDF: gold OA via eScholarship; archive status “pending” — verifier should trigger download and read end-to-end before flipping verified: true
  • Fernandez 2025: closed-access; no local PDF; verified: false must be maintained indefinitely; verifier should note no-fulltext-access
  • verification priority: HIGH — central claim is Singh 2023 lifespan numerics (% extension, cohort n’s, taurine decline %) which are abstract-sourced only; verifier must read Singh 2023 PDF; Fernandez 2025 key finding confirmed from abstract + Crossref, but detailed numerics (cohort n’s, effect sizes) unverified

[2026-05-04] update | taurine.md — Fig. 4A selective-framing critique

External video transcript (molecular biology PhD reviewing Singh 2023) flagged that the paper’s discussion of human correlations selectively emphasizes favorable associations (CRP, obesity, diabetes, glucose ↓ with taurine) while gloss-mentioning the unfavorable associations (LDL, cholesterol, dyslipidemia, AST, GGT, APOB, triglycerides ↑ with taurine in same heatmap).

PDF re-checked directly (Fig. 4A + page 6 text). Critique confirmed: paper text states “For liver- and lipid-related traits such as aspartate aminotransferase (AST) and blood cholesterol, we found positive associations with taurine but negative associations with those of its precursor hypotaurine” — but the next sentence concludes “Association does not establish causation, but these results are consistent with taurine deficiency contributing to human aging,” selectively weighting favorable correlations. The opposite-direction pattern between taurine and hypotaurine is itself a complication that the authors acknowledge but don’t reconcile.

Page updates:

  • EPIC-Norfolk bullet expanded to enumerate both favorable and unfavorable correlation directions explicitly (lines ~77–79 of taurine.md)
  • New caveat added under “Caveats (independent of Fernandez 2025)” section flagging the selective-framing concern
  • Both flagged with #gap/contradictory-evidence
  • No verified flag change required (page already verified: true post-Singh 2023 PDF read; this is a framing/emphasis update, not a numeric correction)

[2026-05-04] update | taurine.md — Sharma 2025 cancer/mTOR concern + Fernandez methodology

External video transcript (Physionic / molecular biology PhD reviewing Fernandez 2025) added two methodological points and one major counterpoint:

  1. Longitudinal vs cross-sectional methodology — explains why Singh 2023 (cross-sectional, different individuals at different ages) and Fernandez 2025 (longitudinal, same individuals over time) can both be technically correct yet disagree. Cross-sectional designs are vulnerable to cohort effects (generation-specific diet/microbiome/smoking confounds) that masquerade as age trajectories. Longitudinal designs are generally given more weight for within-individual aging questions. Added to Fernandez 2025 section as design-quality argument.
  2. Tissue/outcome heterogeneity — Fernandez 2025 reportedly shows higher serum taurine associates with better leg/knee strength in women but NOT with grip strength. Per-outcome specificity argues against uniform “taurine is youth” reading.
  3. Sharma 2025 (Nature, 10.1038/s41586-025-09018-7) — independent line of evidence flagged in passing by the video. This is a major Nature paper (gold/hybrid OA via PMC, 21 citations, FWCI 104.75 — top 1%) showing:
    • Taurine 1.7-fold elevated in leukaemic bone-marrow interstitial fluid
    • Osteolineage-specific CDO1 KO (CDO1^fl/fl Prrx1-cre) extends survival ~13.5% in leukaemia-challenged mice
    • TAUT genetic LOF in patient-derived AML xenografts impairs leukaemia progression
    • TAUT inhibition synergizes with venetoclax against primary human AML
    • CRITICAL mechanism: taurine uptake activates RAG-GTP-dependent mTOR signaling and downstream glycolysis. This is mechanistically OPPOSITE to rapamycin (mTORC1 inhibition → longevity). Singh 2023 reported decreased pRS6P with taurine; Sharma 2025 reports the opposite directionality in stromal contexts. Tissue/cell-type specificity may reconcile but is not addressed in the literature. contradictory-evidence

PDF for Sharma 2025 downloaded (19MB) and pages 1–3 verified directly. Added new section ”## Cancer and mTOR concerns — Sharma 2025 (Nature)” between synthesis and exercise sections. Added Sharma 2025 footnote with full citation + verification scope.

Frontmatter banner upgraded: now flags THREE independent lines of evidence challenging the supplement narrative (Fernandez 2025 biomarker premise, Sharma 2025 leukemogenesis, concurrent HCC/breast cancer reports).

verified flag retained as verified: true with expanded verified-scope documenting Sharma 2025 as partial-page verification.

Implicit stub created: [[studies/sharma-2025-taurine-leukaemogenesis]] (added to R10f study-page deferred queue).

[2026-05-04] update | taurine.md — three additional papers (Tzang 2024 meta-analysis, Chouraki 2017 Framingham, Ito 2023 review)

User supplied three more taurine papers; all verified against local sources to the extent possible.

Tzang 2024 meta-analysis (Nutr Diabetes, 10.1038/s41387-024-00289-z)

  • PDF downloaded (gold OA via PMC); pages 1–5 verified directly
  • 25 RCTs, 1,024 participants; doses 0.5–6 g/day; durations 5–365 d
  • Pooled WMDs (vs control): SBP −3.999 mmHg (p=0.017, I²=85% — high heterogeneity flag), DBP −1.509 mmHg (p=0.002), FBG −5.882 mg/dL (p=0.018), TG −18.315 mg/dL (p<0.001), TC −8.305 mg/dL, LDL ↓ sig, HDL +0.644 mg/dL (ns), BW/BMI ns
  • Critical: directly defuses the EPIC-Norfolk lipid concern from Singh 2023 Fig. 4A — RCT supplementation REDUCES TG/TC/LDL, opposite of cross-sectional positive correlation. Suggests EPIC-Norfolk correlations reflect reverse causation/confounding.
  • Caveats: most trials short, surrogate markers only, 18/25 unclear allocation concealment, subset Taisho-funded, no aging endpoints
  • Added new section ”## Tzang 2024 — meta-analysis of taurine supplementation RCTs for metabolic syndrome” with full effect-size table

Chouraki 2017 Framingham (J Alzheimer’s Dement, 10.1016/j.jalz.2017.04.009)

  • Note: paper authors are Chouraki et al. (NOT “Sun” as initially indexed by archive); first author Vincent Chouraki, senior Sudha Seshadri
  • PDF retrieval failed via all automated paths (Europe PMC backend rejected; PMC HTML wrapper not pdf); abstract verified via PubMed efetch
  • Prospective metabolomics in n=2,067 Framingham Offspring; 217 metabolites; mean 15.6 yr follow-up; 93 incident dementia
  • Taurine HR=0.74 (95% CI 0.60–0.92) for incident dementia — suggestive (not Bonferroni-significant for 217 metabolites)
  • This is the strongest prospective longitudinal observational signal for higher taurine ↔ better brain aging
  • Tension with Fernandez 2025 noted: if taurine doesn’t decline with age, then the Chouraki HR may reflect interindividual variation rather than age-related deficit
  • Added new section ”## Chouraki 2017 — prospective longitudinal taurine and incident dementia (Framingham)“

Ito 2023 review (J Pharmacol Sci, 10.1016/j.jphs.2023.12.006)

  • PDF downloaded (gold OA CC-BY); pages 1–5 verified directly (full review)
  • Heart taurine ~20 mM (~100× plasma); TauT/SLC6A6 mutations cataloged (Shakeel, Ansar, Preising, Garnier); GWAS OR=1.36 for DCM
  • Notable nuances pulled to wiki:
    • Vegetarians do NOT develop DCM despite low taurine intake — explicit refutation of simple “humans are taurine-deficient” framing
    • Ansar Gly399Val homozygotes: blood taurine 6–7 µmol/L vs normal 30–120; rescued by 24 mo oral taurine supplementation
    • Preising Ala78Glu homozygotes: 95% transporter loss, retinal degeneration, but NO cardiomyopathy at 4–11 yr — severe TauT-LOF doesn’t always cause DCM
    • Mixed rodent age-decline literature: F344 rats (liver/kidney/brain decline; heart/muscle unchanged); Massie et al — taurine unchanged in aged C57BL/6 male mice (the strain Singh 2023 used); SD rats unchanged
    • Dawson — 1.5% taurine in drinking water in aged F344 rats restored serum taurine but had NO EFFECT ON LIFESPAN — important null result not prominent in Singh 2023 narrative
    • Verbatim Ito 2023 conclusion: “it is difficult to summarize the effect of age in whole body taurine content, which is likely to be influenced by animal species, strain, sex and age of animal models”
  • Added new section ”## Ito 2023 review — taurine, TauT, dilated cardiomyopathy and ageing”

Page-level changes

  • Synthesis table at top of contradictions section expanded from 2-pillar (interventional vs biomarker) to 5-row picture (mouse lifespan, cross-sectional decline, RCT cardiometabolic benefit, prospective dementia association, mTOR/cancer concern) — better reflects actual evidence landscape
  • Frontmatter banner rewritten: now explicitly characterized as “genuinely mixed — neither pro nor anti”; positive interventional + observational signals acknowledged alongside disputed premises and safety counter-signals
  • Three new footnotes: tzang2024, chouraki2017, ito2023 (full citations + verification scope)
  • Implicit stubs created: studies/tzang-2024-taurine-metabolic-syndrome-meta-analysis, studies/chouraki-2017-framingham-amine-biomarkers-dementia, studies/ito-2024-taurine-deficiency-cardiomyopathy-aging (added to R10f deferred queue)
  • verified flag retained as verified: true; verified-scope expanded to document Tzang/Ito as PDF-verified, Chouraki abstract-only

The page now represents taurine more honestly as a compound with positive RCT cardiometabolic data, suggestive prospective dementia association, disputed age-decline premise, and serious mTOR/cancer mechanistic concern. The “contradictory evidence” framing is preserved but more carefully balanced.

[2026-05-05] update | taurine — mechanistic-theory section added

molecules/compounds/taurine.md

  • Added new section ”## Mechanistic theories — unsubstantiated and untested” between Sharma 2025 (cancer/mTOR) and Exercise/sports sections
  • Section is explicitly flagged as SPECULATION (banner blockquote) and every claim tagged #theory/unsubstantiated; no new empirical claims introduced
  • Three-tier organisation:
    • Tier 1 (mechanistically specific, taurine has known molecular role): mt-tRNA modification (5-taurinomethyluridine on mt-tRNA-Leu/Lys → OXPHOS translation); taurine-chloramine antioxidant flux from HOCl quenching; cardiac/skeletal Ca²⁺ handling via SR + Na⁺/Ca²⁺ exchanger; bile-acid conjugation shifts → FXR/TGR5
    • Tier 2 (downstream cascades through wiki-mapped pathways): Cascade A — mt-tRNA → OXPHOS → ↓mtROS → ↓mtDNA cytosolic leak → ↓cgas-sting → ↓nf-kb → ↓SASP, with parallel ↑NAD⁺ → ↑sirtuin; Cascade B — energy sensing → ampk → ↓mtorC1 → ↑autophagy → ↑mitophagy via pink1-parkin-pathway (with explicit note that Cascade B directly conflicts with Sharma 2025’s mTOR-activating direction in BM niche — tissue-context dependence flagged but unresolved)
    • Tier 3 (alternative explanations not requiring taurine to be the active ingredient): pharmacological supraphysiology at ~6 g/day human-equivalent; mild CR confound from gavage (Singh’s reported ~10% body-weight suppression in T1000 females flagged as directionally CR-consistent; pair-feeding controls absent); microbiome shifts via taurine→H₂S via Bilophila; GABA-A anxiolytic confound on behavioural healthspan readouts
  • Closing synthesis subsection links theories to empirical contradictions: notes that the mt-tRNA deficit-rescue framing has no upstream driver in C57BL/6 if Massie et al’s “tissue taurine unchanged with age” is correct; Tier 3 alternatives collectively predict supraphysiological-dose-only benefits, which is roughly what Tzang 2024’s RCT data shows; most parsimonious read = mix of pharmacological GABA-A/Ca²⁺/antioxidant effects + mild CR confound, with mt-tRNA mechanism contributing tissue- and model-specifically and not reliably reproducible (consistent with Dawson’s null lifespan in F344)
  • New tag introduced (informal, not in CLAUDE.md tag schema): #theory/unsubstantiated — used to mark speculative mechanistic reasoning, distinct from the existing evidence-quality #gap/* tags. CLAUDE.md is silent on speculation tagging; flagging here for user review. If user prefers consolidation under #gap/no-mechanism or a different convention, easy revert.
  • verified flag retained as verified: true: speculation section adds no new source-derived claims, so does not affect existing verified-scope. Banner clearly demarcates speculation from verified content.
  • Origin: user conversation question on whether wiki evidence supports taurine for a 45-yr-old male, followed by request to theorise about mechanisms behind Singh 2023’s positive mouse data and add the theories to the page tagged as unsubstantiated.

[2026-05-05] verify | progeria

phenotypes/progeria.md

  • page: phenotypes/progeria.md
  • sources verified against full PDF (all downloaded this pass unless noted):
    • Eriksson 2003 (10.1038/nature01629) — already in archive; 6 pp read end-to-end
    • Goldman 2004 (10.1073/pnas.0402943101) — downloaded this pass (green OA via Europe PMC); 6 pp read end-to-end
    • Cao 2011 JCI (10.1172/JCI43578) — downloaded this pass (bronze OA via JCI direct); 12 pp read end-to-end
    • Cao 2011 SciTransMed (10.1126/scitranslmed.3002346) — already in archive; 12 pp read end-to-end
    • Gordon 2014 Circulation (10.1161/CIRCULATIONAHA.113.008285) — downloaded this pass (bronze OA via camoufox); 8 pp read end-to-end
    • Gordon 2018 JAMA (10.1001/jama.2018.3264) — downloaded this pass (PMC via Europe PMC); 9 pp read end-to-end
  • sources unverifiable:
    • Scaffidi 2006 (10.1126/science.1127168) — download failed (green OA per metadata, no PMC candidate URLs found); tagged no-fulltext-access in body and footnote
    • De Sandre-Giovannoli 2003 (10.1126/science.1084125) — not_oa; tagged in existing footnote
  • Cao year resolved: Cao 2011 JCI (10.1172/JCI43578) confirmed as the telomere-progerin feedforward paper; “Cao 2007 JCI” was an AI extraction error; the 2011 year in the wiki body was already correct; footnote annotation retained for context
  • corrections made (10):
    1. Survival numbers (CRITICAL): “2.5 years longer (median 14.6 yr treated vs 13.4 yr untreated; log-rank p=0.0013)” — these figures are NOT present in Gordon 2018 JAMA; removed entirely; replaced with actual Gordon 2018 findings: HR 0.12 (95% CI 0.01–0.93; P=.04) ProLon1 analysis; HR 0.23 (95% CI 0.06–0.90; P=.04) combined analysis; untreated natural history mean 14.5 yr, median 14.6 yr; median treatment duration 2.2 yr
    2. “historical controls” → “contemporaneous matched controls” — Gordon 2018 matched treated to untreated by age, sex, continent of residency; not historical controls
    3. Gordon 2014 body text mis-described as “Phase 2 trial (n=25 expanded to 43); primary endpoint rate of weight gain” → corrected to survival analysis (n=43 treated, n=161 untreated; mean survival increase 1.6 yr; P<0.001; HR 0.13, 95% CI 0.04–0.37); Phase 2 clinical outcomes are from Gordon 2012 PNAS, not Gordon 2014 Circulation
    4. Gordon 2014 footnote corrected from Phase 2 weight-gain description to survival analysis description
    5. Age of death in clinical table: “Median 14.5 yr | Death ~90% from MI or stroke [eriksson2003]” → “Mean/median ~14.5–14.6 yr [gordon2018] | at least 90% from progressive atherosclerosis [eriksson2003]; in largest cohort (n=258) heart failure in 79.4% of identified deaths [gordon2018]”; Eriksson 2003 states “average age 13” in own n=20 cohort
    6. Eriksson 2003 footnote: “median age of death from cardiovascular disease” → corrected to “death occurs on average at age 13; at least 90% from progressive atherosclerosis of coronary and cerebrovascular arteries”
    7. Goldman 2004 footnote: removed unattributed “H3K9me3” claim (not in that paper); corrected to actual findings: progressive lobulation (31%→80.9% from P6→P26), EM-documented peripheral heterochromatin loss, lamin-B1 segregation, nuclear pore clustering, early-S-phase PCNA arrest
    8. Goldman 2004 body text: “H3K9me3 and H3K27me3 marks progressively depleted” → “peripheral heterochromatin progressively lost (EM evidence); lamin-B1 segregates in late-passage lobulated nuclei; H3K27me3 reduction is from Cao 2011 SciTransMed”
    9. Hallmark table: “dysfunctional DNA repair via disrupted Rad51/53BP1 scaffolding” removed — not from Goldman 2004; replaced with actual Goldman 2004 findings (replication block, pore clustering)
    10. Limitations section: “~2.5 years on average” survival benefit replaced with accurate HR-based description and 1.6 yr figure from Gordon 2014
  • prevalence frontmatter updated: added “~1 in 20 million living individuals” (from Gordon 2018); schema departure documented (using prevalence: not prevalence-65plus: — justified for pediatric disease)
  • Scaffidi 2006 body claim tagged no-fulltext-access
  • banner removed; verified: true (partial scope)
  • downstream pages to check (main agent):
    • partial-reprogramming — cites Ocampo 2016 in LAKI context; already verified separately; check consistency with progeria.md LAKI description
    • information-theory-of-aging — may reference progerin/normal-aging link (Scaffidi 2006 claim unverified)
    • eriksson-2003-progeria-lmna-mutation — study page if it exists; should reflect corrected age-of-death attribution (Eriksson says “average age 13”; 14.5/14.6 yr from Gordon 2014/2018)
    • cao-2011-rapamycin-progeria — study page (wikilinked in footnote); check consistency with corrected Cao 2011 STM characterization

[2026-05-05] verify | klf4

molecules/proteins/klf4.md

  • page: molecules/proteins/klf4.md
  • sources verified against full PDF:
    • Shields 1996 (10.1074/jbc.271.33.20009) — downloaded this pass (hybrid OA via camoufox); 9 pp read end-to-end
    • Takahashi 2006 (10.1016/j.cell.2006.07.024) — downloaded this pass (bronze OA via camoufox); 14 pp read end-to-end
    • Soufi 2015 (10.1016/j.cell.2015.03.017) — downloaded this pass (bronze OA via camoufox); 14 pp read end-to-end
    • Lu 2020 (10.1038/s41586-020-2975-4) — already in archive; 15 pp read end-to-end
    • Yang 2023 (10.1016/j.cell.2022.12.027) — already in archive; 15 pp read end-to-end
  • sources unverifiable (not_oa): Katz 2002 (10.1242/dev.129.11.2619); Segre 1999 (10.1038/11926) — claims from these sources already appropriately hedged in body; no new tags needed
  • sources not verified (pending/not_oa): Xie 2017 (10.1158/1078-0432.CCR-17-0387); Li 2014 (10.1016/j.juro.2013.08.087)
  • corrections made (8):
    1. hallmarks: frontmatter: removed [[loss-of-proteostasis]] — no body content supports KLF4 as a proteostasis regulator; body correctly describes epigenetic-alterations and cellular-senescence roles only
    2. Structural domains paragraph: zinc-finger residue ranges (430–454, 460–484, 490–512) re-attributed to UniProt O43474 (human) rather than Shields 1996 (which characterized the mouse 483 aa GKLF protein, not the human 513 aa KLF4)
    3. Cell-cycle arrest section: corrected source of cloning — GKLF was cloned from a NIH 3T3 mouse fibroblast cDNA library (growth-arrested cells), not from intestinal epithelial cells; added molecular weight (53 kDa) from source
    4. Soufi 2015 extrapolation table: “crystal structure and binding kinetics” → “EMSA, DNase footprinting, ChIP-seq, and structural modeling; no crystal structure reported” — Soufi 2015 contains no crystallography
    5. Shields 1996 footnote model: “human colon epithelial cells (CaCo-2, HCT-116)” → “NIH 3T3 mouse fibroblasts (functional assays); mouse colon tissue (expression)” — CaCo-2/HCT-116 cells not used in this paper
    6. Lu 2020 footnote model: “5- and 20-month-old; optic nerve crush + glaucoma models” → “C57BL6/J (vision restoration experiments in 3- and 11-month-old mice; systemic safety tested in 5- and 20-month-old mice)” — 5/20 month ages were for safety testing only; vision experiments used 3/11 month mice
    7. Yang 2023 footnote: “assessed to 16 months post-treatment” → “primary phenotypic endpoints assessed ~10 months post-treatment, animals ~14–16 months old” — 16 months was animal age at assessment, not post-treatment duration
    8. Yang 2023 body text: “~57%” → “up to 57%” to match source wording
  • archive updates: Shields 1996, Takahashi 2006, Soufi 2015 footnotes updated from “pending download” to “downloaded”
  • GenAge confirmed: KLF4 not present in GenAge human database (search 2026-05-05); #gap/needs-canonical-id tag resolved; note added to gaps section
  • banner removed; verified: true (partial scope — Katz 2002, Segre 1999, Xie 2017, Li 2014 not verified against PDF; canonical-DB identity fields beyond UniProt not independently re-verified)
  • downstream pages to check (main agent): partial-reprogramming, information-theory-of-aging, lu-2020-osk-vision-restoration, yang-2023-epigenetic-information-loss — these pages should be checked for consistency with the corrected Lu 2020 age-of-subject description (3- and 11-month-old for vision experiments)

[2026-05-05] Round 11 batch summary (compounds + Yamanaka factors + visible-aging phenotypes)

Seeders (11 in parallel, all completed)

Verifiers (11 in parallel, all completed)

Major cross-cutting corrections:

  • Lu 2020 mouse ages (caught by SOX2 + KLF4 verifiers independently): wiki-wide claim that vision experiments used 5- and 20-month-old mice was wrong — those were systemic AAV9 safety-test animals only. Vision experiments used 1- and 3-month-old (optic nerve crush) and 3- and 12-month-old (aging vision restoration). Propagated to partial-reprogramming and information-theory-of-aging footnotes.
  • Yang 2023 timeline (caught by KLF4 verifier): “16 months post-treatment” was animal age at assessment, not post-treatment duration. Correct framing: ~10 months post-treatment, animals ~14-16 mo at assessment. Propagated to partial-reprogramming body + footnote and information-theory-of-aging footnote.
  • Niwa 2000 dosage threshold (caught by OCT4 verifier): “~150% increase” (implies 2.5×) was wrong — paper threshold is ±50% of normal (>1.5× sufficient to drive differentiation). partial-reprogramming body did not duplicate this claim; no propagation needed.
  • Hofmann 2015 lifespan disaggregation (caught by c-MYC verifier): “~15%” was the combined figure; actual data is sex-disaggregated (♀ 20.9% / ♂ 10.7% / combined 15.1%). Critical caveat added: Myc+/- mice show NO improvement in stress pathways (ROS, DNA damage, p21/p16, senescence). hyperfunction-theory cites Myc but does not have lifespan numbers; no propagation needed.
  • Demaria 2014 wound healing mechanism (caught by skin-aging verifier): PDGF-AA “drives myofibroblast differentiation” not “clearance”. Existing cellular-senescence reference uses “promote scar resolution via PDGF-AA secretion” which is acceptable; no edit.
  • Hattori 2004 SCF cell type CRITICAL (caught by skin-aging verifier): solar lentigines SCF source is epidermal keratinocytes, not papillary dermal fibroblasts. Body-wide correction on skin-aging; no other pages cite Hattori 2004.
  • Hamilton 1951 fabricated citation (caught by AA verifier): “Am J Anat 1951;86(3):399-476” does not exist. Real Hamilton papers: Am J Anat 1942;71(3):451-480 (castration mechanism) + Ann N Y Acad Sci 1951;53:708-728 (patterned classification). Corrected on androgenetic-alopecia.
  • Sun 2023 reframing (caught by hair-greying verifier): paper proposes reversible WNT-governed McSC dedifferentiation, not “immobilization”. Section heading + body framing changed.
  • Gordon 2018 progeria survival numbers fabricated (caught by progeria verifier): “2.5 yr longer; median 14.6 vs 13.4 yr; p=0.0013” not in paper. Replaced with actual HR 0.12 (95% CI 0.01-0.93; P=.04) for ProLon1 vs matched controls; HR 0.23 (0.06-0.90; P=.04) for combined analysis.
  • Fabricated p-value in Heathcote 1995 (caught by ear-nose verifier): “p<0.001” not in paper (95% CI 0.17-0.27 mm/yr reported instead); “ear width not significantly associated with age” removed (width never measured in the paper).

Propagation pass (this session)

  • partial-reprogramming line ~86, ~220, ~222: Lu 2020 ages + delivery system corrected; Yang 2023 timeline corrected.
  • information-theory-of-aging line ~198, ~200: same Lu 2020 + Yang 2023 corrections.
  • androgenetic-alopecia line ~212: Nishimura 2002 page range 854-860 → 854-859 (matches hair-greying.md).
  • No other pages required edits (greps for Hofmann, PDGF-AA outside cellular-senescence, Niwa 150%, Cao 2007 returned only the verified pages themselves or log.md history).

ROADMAP updates

Outstanding TODOs surfaced

Graph View

Backlinks