Overview

About This Document

This document explains how to use Neulite, a high-performance simulation environment for biophysical neuron models, and its frontend bionetlite.

_images/767369955-bionetlite_overview.png

Neulite enables models written in Brain Modeling Toolkit (BMTK) to run with minimal modifications. It leverages existing BMTK models while achieving execution in high-performance computing environments.

System Architecture

The Neulite system features a clear separation between frontend and backend components. This separation enables flexible network construction in Python while achieving high-speed simulation execution in C.

  • Backend: Neulite Kernel (C-based)

  • Frontend: Bionetlite Module (Python-based)

This separation allows network construction and actual simulation execution to be independent, enabling flexible development and maintenance.

_images/763493463-bmtk.png

What is Bionetlite?

Bionetlite is a network construction tool compatible with Brain Modeling Toolkit (BMTK). By reusing existing BMTK code with minimal modifications, it automatically supports standard data formats such as SONATA.

  • Network construction and preprocessing tool implemented in Python

  • Performs preprocessing of SWC files (axon correction, DFS sorting)

  • Converts ion channel configuration from JSON to CSV format

  • Automatically generates configuration files for the Neulite kernel

See What is Bionetlite? for details.

What is Neulite Kernel?

Neulite Kernel is a lightweight simulator for executing large-scale biophysical neuron model simulations at high speed. It specializes in the Perisomatic model from the Allen Cell Types Database and runs across a wide range of environments from Raspberry Pi to the Fugaku supercomputer.

  • High-speed simulator written in C (C17)

  • Reads configuration files generated by bionetlite

  • Users can extend functionality in C

See What is Neulite Kernel? for details.

Usage Flow

Preparation: Prepare BMTK network construction code

  1. Network Construction: Execute Python code that imports bionetlite (configuration files for the Neulite kernel are automatically generated)

  2. Simulation Execution: Run simulation with the Neulite kernel

  3. Result Analysis: Analyze the generated data

Next Steps