IsoFind · Base de données isotopiques
Database

Structured and
Certified Isotopic Data

IsoFind structures geochemical isotopic data into two distinct levels: a community database open to scientific contribution, and a professional database certified for forensic and regulatory applications.

10 Supported Elements
Fe · Pb · Zn · Cr Cd · Cu · Sb · Sn · Sr · Se
ISOF Exchange Format
Trust Levels
Community Non-certified data, for comparative and educational use
Professional Certified data, ECDSA signature, traceable provenance
ISOF Format Secure and verifiable exchange between laboratories

Two Databases,
Two Standards of Excellence

Each database serves a distinct purpose: the community database fosters scientific sharing, while the professional database meets forensic and regulatory requirements.

Coming soon
Community Database

Isotopic data compiled from peer-reviewed scientific literature and community contributions. Open access, suitable for teaching, comparison, and exploratory research.

Trust
Non-certified
  • Data from referenced scientific publications
  • Open access, direct integration into IsoFind
  • Comparative, educational, and exploratory use
  • Open to contributions from the scientific community
  • Non-certified: provenance not individually verifiable
Coming soon
Professional Database

Certified isotopic data, cryptographically signed by the issuing laboratories. Designed for forensic, regulatory applications and defensible chains of custody.

Trust
ECDSA Certified
  • ECDSA Signature: each entry signed by the issuing laboratory
  • Full Metadata: protocol, instrument, uncertainties, yields
  • Clear Provenance: laboratory ID and verifiable certificate
  • Versioning: revision history and update traceability
  • Isotopic CRMs: management of certified reference materials
Pre-trained Models

Geochemical ONNX Models

IsoFind publishes machine learning models in ONNX format, trained on 1.7 million geochemical data points. These models focus on water, soil, and sediment geochemistry (not isotopes) and serve as a contextualization foundation for isotopic analyses.

Usable directly from Python, Rust, or any ONNX Runtime compatible environment, without an IsoFind installation.

Geochemical Contextualization
Estimation of regional background levels for trace elements
Anomaly Detection
Abnormal concentrations vs. regional natural backgrounds
IsoFind Nexus Integration
Used natively within the ML/Contextualization module
Speciation Model
Water Geochemistry · 1.7 M points
FormatONNX Runtime ≥ 1.16
TargetChemical speciation in solution
CompatibilityPython · Rust · C++ · Java
Available soon
Adsorption Model
Soil & Sediment Geochemistry · 400 k points
FormatONNX Runtime ≥ 1.16
TargetAdsorption on iron minerals
CompatibilityPython · Rust · C++ · Java
Available soon
Call for Contribution

The Community Database
Needs Your Data

The value of an isotopic database is directly linked to its geographic, elemental, and geological coverage. We are looking for researchers and laboratories ready to share their published data to enrich the community database.

Each contributed dataset is referenced with its authors and original publication. Contribution is entirely voluntary and does not involve any transfer of intellectual property rights over the data.

Data from peer-reviewed publications (any journal)
Georeferenced data with coordinates or geological region
Trace metal isotopes: Sb, Pb, Cu, Zn, Fe, Cd, Tl...
Tabular data with uncertainties and instrumental metadata
Contribution Terms
Data must come from an accessible publication (DOI required)
Contributor must be an author or co-author, or have explicit consent from the authors
Data is integrated with full reference and author attribution
Contribution does not involve any transfer of intellectual property
Accepted formats: Spreadsheet (.xlsx, .csv) or .isof file
Submit data
colin.ferrari@isofind.tech

ISOF Format & Interoperability

Both databases rely on the ISOF format to ensure data portability and verifiability, regardless of the software used.

Open .isof Format

MIT License. Each file embeds data, analytical metadata, and cryptographic proof of integrity. Readable without IsoFind.

Offline Verification

The authenticity and integrity of each entry are verifiable without a network connection or a trusted third party, via the embedded ECDSA signature.

Full Interoperability

Data can be exchanged between laboratories and integrated into any LIMS or third-party system via the Python parser (pip install isof) or Rust.