The ultimate goal of the Blue Brain Facility is to enable neuroscientists to conduct in silico experiments and exploratory studies testing hypotheses, diagnostic tools and treatments for diseases of the brain. This step in the Blue Brain workflow supports such studies by providing facilities for configuring digital reconstructions, setting up experiments and high performance simulations, visualization and analysis, all based on a single, unifying data model.
In silico experiments involve the simulation of biophysical processes at different levels of brain organization, represented by models integrating the data acquired and managed during the first two steps in the Blue Brain workflow, with the mathematical abstractions developed in the third step. The BBP simulation environment
generates numerical solutions for these models, guaranteeing accurate representation of causal relationships across different scales in time and space and allowing users to interact with the model.
The Simulation Environment exploits the computational capabilities of the BBP’s Blue Gene/Q super-computer.
In silico studies require many different kinds of analytical data ranging from the time stamp for a single event -to EEG measurements covering the whole brain. The Analytics Environment leverages the Data Model (see below) to provide a reliable framework for the development of reusable, potentially interactive tools. The framework is designed to support cohosting of simulations and analytics tools on the same machine.
The Visualization Environment provides the interface between the user, modeling tools (Builders) and the simulation and analytics environments, with which it shares a common data model. The Visualization Environment provides users with tools allowing them to display and navigate the structural, transient and analysis data generated by these tools and environments. These tools include virtual instruments allowing users to display in silico brain structures in the same way they appear during experiments with biological samples.
Modelling and simulating the brain requires the integration of structural and functional information spanning many orders of magnitude in space and time. The Blue Brain Data Model provides a unifying “spatial scaffolding” for mathematical models representing different aspects of the brain on different spatial and temporal scales or serving different purposes (e.g. visualization vs. simulation). All Blue Brain modelling, simulation, analytics and visualization are based on instances of the model.