Advancing Bio-Realistic Modeling via the Brain Modeling ToolKit and SONATA Data Format
One of the major goals of the BRAIN Initiative is to distill complex, multi-modal data into predictive frameworks
via theory/modeling. As the planning document "BRAIN 2025: A Scientific Vision" urges, “theory and modeling
should be woven into successive stages of ongoing experiments, enabling bridges to be built from single cells
to connectivity, population dynamics, and behavior.” However, data-driven, bio-realistic modeling is not widely
practiced, in part because the field needs software supporting such complex modeling and standards for model
sharing and reproducibility.
The Allen Institute has developed two powerful tools addressing these needs. One is the Brain Modeling
ToolKit (BMTK) – a software suite for model building and simulation at multiple levels of resolution, from
networks of biophysically detailed neuronal models, to point-neuron networks, to population-statistics
approaches. The other one is the SONATA (Scalable Open Network Architecture TemplAte) data format,
which provides computationally efficient solutions for storing and exchanging data describing all stages of the
modeling workflow (e.g., structure of model networks, configuration of simulations, simulation outputs). These
tools were developed in coordination with many initiatives, such as NEURON, NEST, Neurodata Without
Borders, NeuroML, PyNN, NetPyNE, and the Human Brain Project. As a result, BMTK and SONATA enable
many applications and have generated substantial interest, with many users already employing these tools.
Most recently, BMTK and SONATA were instrumental in integrating diverse data from the Allen Institute and
from the literature into some of the most sophisticated and bio-realistic models of a brain region to date.
We propose to build a comprehensive user support and dissemination platform for BMTK and SONATA and
help integrate these tools into model building and simulation practices in the community. In addition, the Allen
Institute team joins forces with a University of Illinois team that developed a widely used molecular visualization
software VMD. By integrating this software with SONATA, we will leverage its powerful existing capabilities to
offer a free, highly efficient visualization tool for neuroscience modeling. Together, these tools will facilitate free
exchange and reproducibility of models and support sophisticated modeling work – especially in cases of
large-scale biologically realistic models relying on systematic integration of experimental data – for novice and
expert users alike. These contributions will advance the BRAIN Initiative’s priority areas of Theory and Data
Analysis and Integrated Approaches and will strongly facilitate FAIRness (Findability, Accessibility,
Interoperability, and Reuse of digital assets) in neuroscience modeling.
