DOME Registry

DOME Registry manages and provides access to comprehensive Data Optimization Model Evaluation (DOME) annotations for published supervised machine learning studies in biology.

Key Features:

• Structured Annotation System: Implements the DOME recommendations through a structured set of questions covering data handling and processing, optimization, evaluation, and model interpretability.
• External Resource Integration: Uses ORCID, APICURON, and the Data Stewardship Wizard to streamline the annotation process and ensure comprehensive documentation.
• Unique Identification and Scoring: Assigns unique identifiers and DOME scores to publications to foster standardized evaluation of machine learning methods.
• Community Curation: Supports continued growth through community-driven curation efforts.

Scientific Applications:

• Machine Learning Validation: Enhances validation and reproducibility of supervised machine learning research in biology by establishing reporting standards.
• Publication Transparency Assessment: Enables standardized evaluation of whether key details are reported transparently in published machine learning studies.
• Methodological Benchmarking: Allows scientists to compare and assess machine learning methods based on consistent DOME scoring criteria.

Methodology:

The registry collects and organizes annotations according to the DOME recommendations, which specify standards for data handling and processing, optimization, evaluation, and model interpretability in supervised machine learning studies.