Blue Brain building tools aim to build comprehensive digital reconstructions of brain structures at different levels of organization that are compatible with the available experimental data. The strategy of these tools is to identify interdependencies in the experimental data and use them to constrain the reconstruction process. Blue Brain software applications are customizable to be reusable for various brain regions for the benefit of the community.
The Blue Brain Python Optimization Library (BluePyOpt) is an extensible framework for data-driven model parameter optimization that wraps and standardises several existing open-source tools. It simplifies the task of creating and sharing these optimization, and the associated techniques and knowledge. This is achieved by abstracting the optimization and evaluation tasks into various reusable and flexible discrete elements according to established best-practices. Further, BluePyOpt provides methods for setting up both small- and large-scale optimizations on a variety of platforms, ranging from laptops to Linux clusters and cloud-based compute infrastructures. This software is also part of the Human Brain Project Brain Simulation Platform.
Blue Brain NeuroR is a collection of tools to repair morphologies.
NeuroR is distributed as a Python package and uses our python library for analyzing morphologies Blue Brain NeuroM. There are three types of morphology repair with NeuroR: Sanitization – the process ofcurating a morphological file, Cut plane repair – which aims at regrowing parts of morphologies that have been cut out when the cell has been experimentally sliced and Unravelling – the action of “stretching” the cell that has been shrunk because of the dehydration caused by the slicing.
When building a network simulation, biophysically detailed electrical models (e-models) need to be tested for every morphology that is possibly used in the circuit. E-models can e.g. be obtained using BluePyOpt by data-driven model parameter optimization. Developing e-models can take a lot of time and computing resources. Therefore, these models are not reoptimized for every morphology in the network. Instead we want to test if an existing e-model matches that particular morphology `well enough’. This process is called Cell Model Management (MM). It takes as input a morphology release, a circuit recipe and a set of e-models with some extra information. Next, it finds all possible (morphology, e-model)-combinations (me-combos) based on e-type, m-type, and layer as described by the circuit recipe, and calculates the scores for every combination. Finally, it writes out the resulting accepted me-combos to a database, and produces a report with information on the number of matches – BluePyMM. This software is also part of the Human Brain Project Brain Simulation Platform.