NeuroML-DB

NeuroML-DB provides a repository of computational neuroscience models translated into the NeuroML model description language to enable sharing, evaluation, and reuse of ion channel, cell, and network models.

Key Features:

• Model Repository: Hosts over 1,500 published models encompassing ion channels, cells, and networks translated into NeuroML.
• Integration with Other Databases: Maintains reciprocal links to ModelDB and Open Source Brain and links to original publications in PubMed.
• Search and Analysis Capabilities: Supports search functionality including integration with the Neuroscience Information Framework (NIF) and computational assessment of electrophysiology, morphology, and computational complexity.
• Modular Model Description: Uses NeuroML as an intermediate language to translate models into other simulator formats, promoting interoperability across simulation software.
• Automated Characterization and Analysis: Enables automated characterization of standardized NeuroML models and database-scale analyses that cluster model-derived features, exemplified by identification of a tetrahedral structure among cell model properties.

Scientific Applications:

• Model reuse for simulation studies: Reuse published ion channel, cell, and network models for simulation-based investigations of neural dynamics.
• Comparative analysis of model properties: Compare electrophysiology, morphology, and computational complexity across models to identify relationships and clusters.
• Discovery of organizational principles: Support database-scale analyses to reveal structures in model property space, such as the reported tetrahedral arrangement of cell model properties.

Methodology:

Published models are translated into the NeuroML model description language; models are linked to ModelDB, Open Source Brain, and PubMed; search is integrated with the Neuroscience Information Framework (NIF); automated characterization quantifies electrophysiology, morphology, and computational complexity; translations to other simulator formats and database-scale clustering analyses (identifying a tetrahedral structure among cell model properties) are performed.