The Virtual Brain Ontology: A Digital Knowledge Framework for Reproducible Brain Network Modeling.

Area of Science:

• Computational neuroscience
• Systems neuroscience
• Neuroinformatics

Background:

• Computational models of brain network dynamics provide insights into brain function and disease.
• Lack of standardization in reporting models, parameters, and code hinders reproducibility and comparability of results.
• Existing shared code is often poorly documented, non-standardized, or not executable.

Purpose of the Study:

• To introduce The Virtual Brain Ontology (TVB-O), a semantic knowledge base and minimal metadata standard.
• To simplify the description, execution, and sharing of computational brain network simulations.
• To enhance the reproducibility, comparability, and portability of computational neuroscience models.

Main Methods:

• Developed a common vocabulary and ontology for simulating brain network dynamics.
• Defined a minimal metadata specification for experimental reproducibility.
• Created a curated database of published models, brain networks, and study configurations.
• Implemented a Python toolbox generating executable code for various simulation platforms (e.g., The Virtual Brain, Jax, Julia).

Main Results:

• TVB-O provides a standardized framework for describing and sharing computational neuroscience models.
• The system enables the generation of FAIR metadata and provenance-aware reports.
• Executable code can be generated for multiple simulation platforms, enhancing model portability.
• Assumptions are made explicit, linking models to biomedical knowledge and observation pathways.

Conclusions:

• TVB-O reduces technical barriers in computational neuroscience, promoting broader access.
• It establishes a foundation for transparent and shareable "digital brain twins."
• The ontology enhances the integration of computational models with clinical pipelines and large-scale data resources.