Abstract
Brian 2 is a fundamental rewrite of the popular Brian [1,2] simulator for spiking neural networks. It is written in the Python programming language and focuses on simplicity and extensibility: neuronal and synaptic models can be described using mathematical formulae and with the use of physical units [3]. Brian has been traditionally used to simulate networks of single-compartment neurons; complex morphologies, modeled as multi-compartment neurons, were not supported. We have recently added support for such models: a neuron's morphology can be either created from a file (in the SWC format [4]) or be created iteratively in the simulation script itself, using a convenient syntax. The flexibility of the Brian simulator applies to multi-compartment models in the same way as to single-compartment models: the model dynamics are governed by arbitrary user-provided differential equations, models can be flexibly parameterized over all the compartments and all variables describing the internal state of the model (voltages, currents , conductances, etc) are accessible to the user. The simulation of multi-compartment models is more complex than that of non-electrically coupled single-compartment models: in addition to the numerical integration of the differential equations governing the change of the membrane potential in a single compartment, the currents flowing across neighboring compartments have to be taken into account. To this end, Brian employs a domain-decomposition method which corrects an initial solution valid for uncoupled compartments, taking the actual coupling into account. This method has the advantage that it is easy to parallelize, since most of computations are performed independently for each compartment. We demonstrate the advantage of this by parallelizing the computation over multiple processor cores using the OpenMP framework. For the future, we are planning to further accelerate the simulation speed by making use of the massively-parallel processing capabilities of graphical processing units. With this addition, we have significantly enlarged the range of research questions that can be investigated using Brian. Brian is made available under a free software license and all development takes place in public code repositories.
Highlights
Brian 2 is a fundamental rewrite of the popular Brian [1,2] simulator for spiking neural networks
The flexibility of the Brian simulator applies to multi-compartment models in the same way as to single-compartment models: the model dynamics are governed by arbitrary user-provided differential equations, models can be flexibly parameterized over all the compartments and all variables describing the internal state of the model are accessible to the user
The simulation of multi-compartment models is more complex than that of non-electrically coupled singlecompartment models: in addition to the numerical integration of the differential equations governing the change of the membrane potential in a single compartment, the currents flowing across neighboring compartments have to be taken into account
Summary
Brian 2 is a fundamental rewrite of the popular Brian [1,2] simulator for spiking neural networks. Brian has been traditionally used to simulate networks of single-compartment neurons; complex morphologies, modeled as multi-compartment neurons, were not supported.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
