Built as a shared layer
- API-first: Symfony + API Platform backend
- Osoma: Vite + React frontend (self-hostable)
- Sits above ELN, LIMS, and AMS — no workflow disruption
- Keycloak OIDC auth, Docker Compose deployment
Open Source
Research-grade FAIR metadata infrastructure for preclinical research
Metadatapp is an open-source metadata management platform for biomedical and laboratory research. Symfony + API Platform backend. Osoma (Vite + React) frontend. JSON-LD, RO-Crate, and ontology-based data modelling — FAIR by design, AI-ready from the start.
✦FAIR by design✦JSON-LD / RO-Crate✦Self-hostable AGPL-3.0
Open standards by design
- JSON-LD and RO-Crate for machine-readable exports
- Ontology-based modelling for semantic interoperability
- ARRIVE-aligned experimental descriptions
- AI-ready: clean, structured research data outputs
What Metadatapp does
Metadatapp is an open-source (AGPL-3.0) metadata management platform combining a Symfony + API Platform backend with the Osoma Vite + React frontend. It is designed as a shared metadata layer — not a replacement for existing lab tools.
FAIR metadata layer (semantic, ontology-driven)
Structures experimental metadata using controlled vocabularies, JSON-LD contexts, and RO-Crate packaging. Every subject, protocol, assay, and outcome is captured with provenance and semantic links — so data stays reusable without manual cleanup.
Interoperability layer (ELN, LIMS, lab systems)
Server-side ConnectedApps integrations connect the tools labs already use — elabFTW, LIMS, AMS, ORCID, ROR — without exposing credentials to the frontend. Metadata flows in, gets validated, and flows out in open formats.
AI-assisted workflows
Optional server-side AI assistant (OpenAI / Anthropic, disabled by default). Credentials managed securely via Osoma's AI Providers settings — never hard-coded.
Who benefits?
Academic labs
Self-hostable, open source, and FAIR-compliant. Structure experiments, export to repositories (OSF, Zenodo), and make your data reusable for future collaborators.
Platform & data teams
REST API with JSON-LD contexts, RBAC, audit logs, Keycloak OIDC, and Docker Compose deployment. Extensible and community-driven.
Shared facilities
Standardised metadata across instruments, studies, and local workflows. Clean, structured outputs reduce manual rework and improve handover quality.
Translational teams
Reuse structured experimental descriptions across projects, cohorts, and repositories without losing scientific context.
Data governance & security
Access & provenance
- Role-based access control (RBAC) via Keycloak OIDC
- Audit logs and provenance capture across all studies
- Validation rules enforced at API layer (API Platform)
Deployment
- Self-hostable via Docker Compose
- Castor task runner for local setup and CI tasks
- Separate auth domain (auth.metadatapp.test / Keycloak)
- Development fixtures and sample data included
Seamless Integrations
elabFTW (ELN)
Direct integration with elabFTW for experiment capture and metadata extraction.
LIMS / AMS Systems
Connect with laboratory and colony management systems via server-side ConnectedApps adapters.
ORCID / ROR
Use persistent researcher and institution identifiers.
AI Providers
Optional OpenAI or Anthropic integration — disabled by default, credentials stored server-side and never exposed to the browser.
Repositories
Export to OSF, Zenodo, and other research repositories via RO-Crate and JSON-LD outputs.
Extend the platform
AGPL-3.0 source code. Add ConnectedApps adapters, custom ontologies, or new frontend modules via the Osoma React app.
Standards & data formats
FAIR Principles
- Findable: rich metadata and search indexing
- Accessible: open API with OIDC auth
- Interoperable: JSON-LD contexts + RO-Crate packaging
- Reusable: provenance, versioning, and ARRIVE alignment. Learn more in FAIR-by-Design.
3Rs Alignment
- Reduction: structured data enables reuse, fewer repeats
- Refinement: richer metadata captures welfare-relevant variables
- Replacement: AI-ready datasets accelerate in-silico modelling
- Better data is the most direct path to animal welfare
FAIR + 3Rs together
FAIR data and 3Rs are inseparable: structured, reusable metadata directly reduces animal use. Better data is better welfare.
Data outputs & exports
Architecture at a glance
│
▼
API Platform (Symfony / PHP 8.4)
│
┌────────┴────────┐
▼ ▼
PostgreSQL ConnectedApps (metadata) (ELN, LIMS, AMS, AI)
│
▼
Keycloak (OIDC auth)
Data flow
→ Validate (JSON-LD + ontology)
→ Store (PostgreSQL + provenance)
→ Export (RO-Crate / JSON-LD / CSV)
→ Reuse (repositories, AI pipelines, collaborators)