Research synthesis, not medical advice — and a work in progress
This is a living, AI-maintained knowledge base, not clinical guidance. Coverage is uneven: some pages are thin, unverified, or absent entirely — the absence of a page or a claim is not evidence of absence. Nothing here substitutes for a clinician.
This is a knowledge base on the cellular and molecular mechanisms of human aging — a densely cross-linked set of atomic pages (one per protein, pathway, compound, process, hallmark, phenotype, biomarker, study, hypothesis, and model organism) built to be operated by an AI agent, and browsable by humans. Every claim cites a primary source and every page is flagged unverified (AI-drafted) or verified against its primary-source PDF so you know how far to trust it. It doesn’t duplicate UniProt, KEGG, PubChem, or PubMed — it cites them by ID and links out; the value here is the aging-specific synthesis, the cross-links, and the tracked gaps, not the raw data.
Source, schema & contributing: the full project — including the operating manual and the seeder/verifier agents — lives on GitHub.
Start here
New here and just want to read? While the intent is to point a local agent like Claude Code at the GitHub Repository, you’re more than welcome to manually browse around here. A few good places to begin:
The two maps — the same biology, sliced two ways:
- hallmarks-of-aging — the academic standard, organized by molecular damage type
- sens-damage-categories — SENS’s intervention-oriented framing
Core mechanisms:
- cellular-senescence — “zombie” cells that accumulate with age and drive inflammation
- autophagy — the cell’s recycling system, and why it falters
- mTOR & IGF-1 — the nutrient-sensing pathways at the center of aging
Interventions:
- caloric restriction — the most replicated lifespan intervention in biology
- senolytics — drugs that clear senescent cells (dasatinib + quercetin, fisetin)
- rapamycin & metformin — the two most-studied geroprotective drugs
Measuring & comparing aging:
- Horvath clock — the original DNA-methylation “epigenetic clock”
- naked mole-rat — a rodent that barely ages; what it teaches us
- information theory of aging — a leading theory of why we age
Or use search, follow the graph, or browse the full catalog below. For how the wiki is built and how to run an agent in it, see the project README.
Browse by entity
Truth lives on atomic pages — one per entity. Each link opens that folder:
Molecules & pathways
- Proteins — genes and gene products
- Compounds — single chemical entities: drugs, small molecules, single-molecule supplements (creatine, CoQ10); each holds its own mechanism + efficacy
- Metabolites — endogenous small-molecule intermediates
- Pathways — signaling pathways
Cells, tissues & systems
- Processes — autophagy, apoptosis, senescence, …
- Cell types — stem, immune, and post-mitotic cell populations
- Tissues — organs and tissues
- Organ systems — the 11 body-system overlays
- Microbes — gut / skin taxa behind the dysbiosis hallmark
Phenotypes, measures & theories
- Hallmarks — the 12 hallmarks of aging (plus the proposed disabled-adaptive-immunity candidate)
- Phenotypes — sarcopenia, frailty, inflammaging, …
- Biomarkers — biological-age clocks (epigenetic / proteomic / composite)
- Hypotheses — contested or open theories of why we age
Interventions & evidence
- Interventions — modalities and multi-compound classes: lifestyle, dietary, pharmacological classes (e.g. senolytics), gene / cell / blood-product therapies; combination & regimen products (e.g. multivitamins) live here
- Model organisms — mouse, worm, fly, killifish, naked mole-rat, human, …
- Studies — primary sources extracted from
- Methods — lab / analytical techniques and their recurring caveats
Maps & frameworks
Navigational overlays that re-slice the atomic layer — they link to entity pages, never duplicate them. Full set in the frameworks folder; the essential ones:
The two damage maps (co-equal)
- hallmarks-of-aging — LĂłpez-OtĂn et al. (12 hallmarks, 2023 update)
- sens-damage-categories — SENS Research Foundation (7 damage categories)
Causal ordering
- hallmark-causality-graph — wiki-internal upstream→downstream ordering of the hallmarks
- causal-graph-data — companion edge-list with evidence grades + per-edge page anchors
Slice the atomic layer by…
- by-organ-system — anatomy (the 11 body systems)
- sex-differences-in-aging — biological sex (reproductive-endocrine axes, sexed mechanisms, disease sex-bias)
- biological-age-measurement — biomarkers / aging clocks
Intervention overlays
- interventions-by-hallmark — evidence matrix by hallmark
- interventions-by-modality — by delivery (drug / cell / dietary / gene / blood-product)
- intervention-classes — mechanism-class taxonomy
- intervention-by-target-immunogenicity — immunogenicity-tier filter
Project & maintenance
- Knowledge gaps — README aggregates open questions by category; planned-coverage holds proactive coverage intent.
- Personal-application layer — apply the research to one person (baseline, plans, labs, tracking). Real data lives in a private, git-ignored
protocols/repo; protocols-template is a fake-data skeleton + onboarding guide. Privacy model in CLAUDE § “Privacy posture”. - Schema & operations — CLAUDE (schema + operating manual) · lint-pass (periodic health check) · acronyms (glossary) · log (change history).
- SOP playbooks (data-source how-tos) — full set in sops/; e.g. retrieving-papers, extracting-evidence, finding-pathway-data, finding-protein-data, finding-compound-data, finding-aging-specific, adding-a-claim.