Related Experiment Video
Updated: Feb 25, 2026
05:55
Modeling the Functional Network for Spatial Navigation in the Human Brain
Published on: October 13, 2023
1.6K
Software for Brain Network Simulations: A Comparative Study
Ruben A Tikidji-Hamburyan1, Vikram Narayana1, Zeki Bozkus2
1School of Engineering and Applied Science, George Washington University, Washington, DC, United States.
Frontiers in Neuroinformatics
|August 5, 2017
Summary
Evaluating computational neuroscience simulators NEURON, GENESIS, BRIAN, and NEST reveals performance biases. While all aim for broad model support and efficiency, NEST excels in high-performance computing, BRIAN offers concise coding, and NEURON suits detailed models.
Area of Science:
- Computational neuroscience
- Neuroscience software development
- Brain network modeling
Background:
- Numerical simulations are vital for understanding brain function.
- Numerous software packages and simulators exist for computational neuroscience research.
- Popular simulators include NEURON, GENESIS, BRIAN, and NEST.
Purpose of the Study:
- To independently evaluate four leading brain network simulators: NEURON, GENESIS, BRIAN, and NEST.
- To compare their model support, computational architecture, efficiency, high-performance computing capabilities, user support, and usage frequency.
- To identify simulator biases for different types of brain network models.
Main Methods:
- Comparative analysis of simulator features and performance.
- Evaluation based on model support range, computational architecture, and efficiency.
- Assessment of high-performance computing amenability (clusters, multicore).
- User support and usage frequency analysis.
- Two case studies: large-scale simplified networks and small-scale detailed networks.
Main Results:
- All simulators compile intensive procedures into binary code for performance.
- NEST offers seamless high-performance computing integration; NEURON requires modifications; BRIAN has limited parallelization.
- BRIAN provides the most concise language for model implementation.
- NEST is favored for large network models; NEURON is better for detailed models.
- Simulators exhibit computational performance biases toward specific model types.
Conclusions:
- Brain network simulators show inherent biases in computational performance and suitability for different model types.
- NEST excels in large-scale simulations and high-performance computing.
- NEURON is optimal for detailed, small-scale network simulations.
- BRIAN offers coding simplicity but has limitations in parallelization.
