Projects

FaceBase

FaceBase is an NIH-funded, open community hub that curates and serves more than 1,100 FAIR datasets in craniofacial and dental biology. Powered by the DERIVA platform, the repository supports self-curation by researchers, integrates interactive imaging and single-cell viewers, and assigns DOIs for every dataset to ensure citation and reuse.

Protein Data Bank - Integrative and Hybrid Methods (PDB-IHM))

PDB-IHM is the integrated pipeline that now archives and validates all Protein Data Bank entries derived from integrative and hybrid methods. Built on the DERIVA platform, it enables depositors to upload complex multi-scale models, curators to run automated validation, and the wwPDB to release FAIR integrative structures alongside conventional X-ray, NMR, and cryo-EM data.

EyeAI

EyeAI combines DERIVA-ML with GPU-based notebooks to create a fully traceable pipeline for glaucoma detection from fundus photographs, spanning data ingest, feature extraction, model training, and clinical review. By treating each dataset, feature set, workflow, and execution as a first-class, versioned artefact, the project shows how DERIVA-ML delivers continuous FAIRness for real-world medical AI.

Synapse

The “Mapping the Dynamic Synaptome” project uses DERIVA to capture, curate, and share single-synapse–resolution imaging of larval zebrafish brains, revealing region-specific synapse gain and loss during memory formation. All raw and processed data are available through the Synapse repository powered by DERIVA, with persistent DOIs for every figure and dataset.

Common Fund Data Ecosystem

The NIH Common Fund Data Ecosystem (CFDE) used DERIVA to prototype its first cross-program metadata catalog and search portal, harmonising descriptions from 11 Data Coordination Centers (DCCs) into a single schema (C2M2) and ingest workflow.

Pancreatic β-Cell Consortium

The Pancreatic β-Cell Consortium (PBCC) uses DERIVA to archive, curate, and share heterogeneous data on β-cell biology and whole-cell modelling, ranging from soft-X-ray tomography and mass-spectrometry to computational models. DERIVA’s model-driven interface lets bench scientists, not DBAs, evolve the schema as new assays appear.