AI-Driven Metadata Curation and Harmonization
Improving the findability and reuse of pathogen datasets by completing, normalizing, and connecting sparse BioProject, BioSample, SRA, and BRC metadata.
Goal
Improve the findability and reuse of pathogen datasets by completing, normalizing, and connecting sparse BioProject, BioSample, SRA, and BRC metadata.
Three-Day MVP
Select approximately 5,000 BioProjects or 10,000 pathogen records. Normalize host, isolation source, geography, collection date, disease, organism, sequencing strategy, and funding fields. Use linked publications and related records as evidence for proposed corrections.
Create a graph linking studies, samples, sequences, publications, organisms, diseases, repositories, and NIAID programs.
Model and Evaluation
Train or calibrate ontology-linking and metadata-normalization components. Hide known metadata fields and measure exact match, hierarchical ontology match, confidence calibration, and unsupported-completion rate.
Team
Team assignments are being finalized ahead of the codeathon. Participants can review project teams, and request reassignment, in the shared participant spreadsheet circulated by the organizing team.