Home

OpenLongevity — Wiki

Author: Ciprian Ștefan Pleșca — Founder, Project Creator, Lead Maintainer and Principal Author Repository: Ciprian-LocalPulse/OpenLongevityLab License: Apache-2.0 Status: Research prototype (package version 0.3.0)

Research use only. OpenLongevity does not diagnose, prevent, treat, or cure disease, and it does not establish that any intervention extends human lifespan.

What this project is

OpenLongevity is an open-source computational research infrastructure for organizing, scoring, and navigating the scientific literature on biological aging. It is built around one central idea: every piece of evidence — a publication, an experimental finding, a synthetic demonstration — must carry its provenance and its uncertainty with it, rather than being flattened into a single unqualified claim.

The repository combines four kinds of work that this wiki treats as distinct engineering surfaces, deliberately not blurred together:

1. A typed scientific core (src/openlongevity) that models evidence, grades it, detects research gaps, and exposes it through a versioned API.
2. Read-only literature adapters for PubMed, Europe PMC, OpenAlex, Crossref, and ClinicalTrials.gov, each normalizing external records into a common Publication shape with full source attribution.
3. A PostgreSQL persistence layer with immutable revision history for ingested publications, kept structurally separate from the synthetic evidence fixtures used for demonstration.
4. Supporting surfaces: a biomarker catalog, a small in-memory evidence graph, statistical analysis utilities (biological-age regression, Kaplan–Meier survival, pathway enrichment, multi-omics joins), a TypeScript dashboard shell, and an optional Rust performance kernel.

Why the separation matters

A recurring theme in this codebase — and in this wiki — is that software behavior is not scientific validation. A successful call to /api/v1/evidence demonstrates that a fixture was returned correctly; it does not demonstrate that a biological claim has been reviewed. The project's own documentation is explicit about this, and the architecture reflects it structurally: fixture-backed endpoints are tagged "mode": "fixture-only", persisted publication endpoints are tagged "mode": "persisted", and every relevant response carries a disclaimer string.

System orientation

flowchart TB
    subgraph External["External scientific sources"]
        PM[PubMed / NCBI E-utilities]
        EPMC[Europe PMC]
        OA[OpenAlex]
        CR[Crossref]
        CT[ClinicalTrials.gov]
    end

    subgraph Core["openlongevity core (Python)"]
        PROV[Provider adapters]
        REPO[PublicationRepository]
        EVID[EvidenceEngine]
        GAPS[ResearchGapDetector]
        GRAPH[EvidenceGraph]
        ANLY[Analysis utilities]
    end

    subgraph Storage["Persistence"]
        PG[(PostgreSQL)]
    end

    subgraph Interfaces["Interfaces"]
        API[FastAPI service]
        CLI[CLI]
        WEB[TypeScript dashboard shell]
    end

    PM --> PROV
    EPMC --> PROV
    OA --> PROV
    CR --> PROV
    CT --> PROV
    PROV --> REPO
    REPO --> PG
    EVID -.synthetic fixtures.-> API
    GAPS -.synthetic fixtures.-> API
    GRAPH -.illustrative.-> API
    REPO --> API
    API --> WEB
    API --> CLI
    ANLY -. standalone utilities .-> Core

Loading

How to read this wiki

This documentation is organized by concern rather than by file, so that each page can be read as a self-contained academic note. Start here, then follow the path that matches what you need:

Page	Purpose
Architecture	Component boundaries, module responsibilities, request lifecycle
Domain-Model	The typed core: EvidenceRecord, Publication, ResearchGap, enumerations and invariants
Evidence-Grading-Engine	The A–G evidence hierarchy and the navigation-score formula, with worked examples
Research-Gap-Detection	Heuristics used to surface translational, replication, and concentration gaps
Literature-Provider-Adapters	PubMed, Europe PMC, OpenAlex, Crossref, ClinicalTrials.gov — contracts and failure handling
Persistence-and-Data-Governance	PostgreSQL schema, immutable revision history, checksum-based change detection
API-Reference	Every route, its authentication requirements, and its response contract
Analysis-Toolkit	Biological-age regression, Kaplan–Meier survival estimation, pathway enrichment, multi-omics integration
Scientific-Methodology-and-Limitations	What the system can and cannot claim, mapped against its own stated limitations
Contribution-and-Governance-Model	Review standards, data governance, security posture, and how to contribute

Quickstart (for orientation only)

python -m venv .venv
pip install -e ".[dev,api,db]"
pytest
openlongevity evidence "cellular senescence"
uvicorn 'openlongevity.api:create_app' --factory --reload

Synthetic records returned by this quickstart are software fixtures, not scientific findings. See Scientific-Methodology-and-Limitations before citing anything the running system returns.

Attribution

This project, its architecture, and its academic framing are authored and maintained by Ciprian Ștefan Pleșca. See AUTHORS.md and CITATION.cff in the repository root for citation metadata.