NAMO: New Approach Methodology Ontology and Schema
Overview
The New Approach Methodology Ontology and Schema (NAMO) is a comprehensive data framework designed to standardize and integrate information about alternative methods to animal testing in biomedical research. Built on LinkML, NAMO provides a common vocabulary and data structure for describing diverse model systems, from cellular cultures and organ-on-chip devices to computational models and digital twins.
Why NAMO Matters
The Challenge
The biomedical research landscape is rapidly evolving with innovative alternatives to traditional animal models:
- Organ-on-chip devices that mimic human tissue physiology
- 3D organoids that recapitulate organ development and disease
- In silico models including AI/ML systems and physiologically-based pharmacokinetic (PBPK) models
- Advanced cellular systems with complex co-cultures and microenvironments
However, these New Approach Methodologies (NAMs) exist in isolated silos with: - Inconsistent terminology across research groups and disciplines - Fragmented data that cannot be easily compared or integrated - Limited discoverability of relevant models for specific research questions - Unclear relationships between model capabilities and human disease contexts
The Solution
NAMO addresses these challenges by providing:
- Unified Data Model: A comprehensive schema that captures the essential characteristics of all NAM types under a common framework
- Ontology Integration: Deep integration with established biomedical ontologies (Cell Ontology, UBERON, MONDO) ensuring semantic interoperability
- Rich Metadata: Detailed capture of model specifications, concordance metrics, and validation data
- Cross-Platform Compatibility: Generation of multiple data formats (JSON Schema, OWL, Python models) for diverse technical ecosystems
Core Applications
Integrated Database Development
NAMO serves as the foundational schema for building comprehensive databases that:
- Catalog NAM systems across research institutions and companies
- Enable cross-model comparisons through standardized metadata
- Support model discovery via semantic search and filtering
- Track validation evidence and concordance with human biology
- Facilitate regulatory acceptance through structured documentation
Research Integration Platforms
Organizations can use NAMO to:
- Harmonize data from multiple model types and sources
- Build knowledge graphs connecting models to diseases, pathways, and phenotypes
- Support AI/ML applications with structured training data
- Enable meta-analyses across different experimental approaches
- Create recommendation systems for optimal model selection
Regulatory Documentation
NAMO supports regulatory science by:
- Standardizing evidence packages for model validation
- Tracking concordance metrics between models and human biology
- Documenting use contexts and limitations transparently
- Enabling systematic reviews of model performance across studies
Key Features
Hierarchical Model Organization
NAMO organizes model systems into three main categories:
- Cellular Systems: 2D/3D cultures, organoids, co-cultures, and cell line models
- Microphysiological Systems: Organ-on-chip and tissue-on-chip devices
- In Silico Models: QSAR, PBPK, digital twins, and machine learning models
Each category captures domain-specific attributes while sharing common metadata patterns.
Rich Concordance Assessment
The schema includes sophisticated frameworks for documenting how well models recapitulate human biology:
- Molecular similarity with gene expression correlation and pathway conservation
- Phenotype overlap using standardized phenotype ontologies
- Functional parity through quantitative assay comparisons
- Cell type coverage with single-cell resolution
- Reproducibility metrics across batches and laboratories
Ontology-Driven Interoperability
NAMO leverages established biomedical ontologies:
- Cell Ontology (CL) for standardized cell type annotations
- UBERON for anatomical structure references
- MONDO for disease associations
- Gene Ontology (GO) for molecular function and pathway annotations
This enables seamless integration with existing biomedical databases and knowledge graphs.
Flexible Validation Framework
The schema supports diverse types of validation evidence:
- Literature references with DOI/PMID tracking
- Experimental assays with methodology and performance metrics
- Omics data integration including transcriptomics and proteomics
- Quality control metrics with pass/fail status tracking
- Cross-validation results for computational models
Technical Implementation
Multi-Format Generation
From a single LinkML schema definition, NAMO automatically generates:
- Python dataclasses and Pydantic models for programmatic access
- JSON Schema for web applications and API validation
- OWL ontology for semantic web applications
- GraphQL schemas for modern API development
- Documentation with cross-referenced class and property descriptions
Extensible Architecture
The schema is designed for evolution:
- Abstract base classes enable addition of new model types
- Flexible slot definitions accommodate emerging measurement modalities
- Enum extensibility supports new controlled vocabularies
- Mixin patterns allow cross-cutting concerns like provenance tracking
Validation and Quality Control
NAMO includes comprehensive validation mechanisms:
- Schema validation ensures data integrity and completeness
- Ontology term validation verifies proper vocabulary usage
- Cross-reference checking maintains referential integrity
- Example validation through curated test cases
Getting Started
For Database Developers
Use NAMO as your foundational schema for NAM data integration:
from namo.datamodel import Dataset, Organoid, QSARModel
# Create a dataset with multiple model types
dataset = Dataset(
model_systems=[
Organoid(
id="org_001",
name="Brain Organoid Model",
organ_modeled={"id": "UBERON:0000955", "name": "brain"}
),
QSARModel(
id="qsar_001",
name="Hepatotoxicity Predictor",
activity_endpoint="DILI classification"
)
]
)
For Data Integration
Leverage NAMO's ontology mappings for semantic interoperability:
model_system:
id: "organoid:001"
type: "Organoid"
cell_types:
- id: "CL:0000540" # neuron
name: "neuron"
- id: "CL:0000127" # astrocyte
name: "astrocyte"
models:
- biological_system_modeled: "alzheimers_disease"
concordance:
molecular_similarity: "0.85"
phenotype_overlap: "High concordance with patient tissue"
For Computational Applications
Build on NAMO's structured data for AI/ML applications:
- Feature extraction from standardized model metadata
- Similarity scoring using ontology-based distance metrics
- Recommendation systems for model selection
- Knowledge graph embeddings for novel association discovery
Community and Standards
NAMO is developed as an open standard with broad community input:
- Open source development on GitHub with transparent governance
- Community-driven extension through working groups and feedback
- Standards-compliant with FAIR data principles and biomedical ontology practices
- Interoperable with existing databases like Monarch Initiative and EBI resources
The schema serves as a bridge between the traditional model organism research community and emerging NAM technologies, providing a path for integrated translational research platforms.
Future Directions
NAMO continues to evolve to meet emerging needs:
- Expanded model types including digital twins and multi-organ systems
- Enhanced concordance metrics with quantitative benchmarking frameworks
- Regulatory alignment with guidelines from FDA, EMA, and OECD
- AI integration with structured training data for model recommendation systems
- Real-time data capture from connected laboratory instruments and devices
By providing a robust, extensible, and community-driven foundation, NAMO enables the next generation of integrated biomedical research platforms that harness the full potential of alternative methodologies for human health advancement.
Documentation
- Complete Schema Documentation - Auto-generated documentation for all classes, slots, and enumerations
- Use Cases and Examples - Practical applications and implementation patterns