bionetlite Setup

This page explains the setup procedure for bionetlite, a Python package for network construction.

Prerequisites

The following software is required:

  • Python 3.7 or higher

Setup Procedure

Step 1: Create Virtual Environment (Recommended)

It is recommended to create a Python virtual environment:

# Create virtual environment
python -m venv neulite_env

# Activate virtual environment (Linux/Mac)
source neulite_env/bin/activate

Step 2: Install BMTK and NEURON

bionetlite depends on Brain Modeling Toolkit (BMTK) and NEURON.

Installing BMTK:

pip install bmtk

Installing NEURON:

Install NEURON (biophysical neuron simulator):

pip install neuron

Note

For more information on BMTK and NEURON, refer to their official sites: BMTK / NEURON

Step 3: Install Required Packages

Installing scikit-learn:

bionetlite requires scikit-learn because it uses PCA (Principal Component Analysis) for axon orientation estimation:

pip install scikit-learn

Step 4: Clone the neulite Repository

Clone the neulite repository:

git clone https://github.com/neulite/neulite.git

The repository contains the following:

  • bionetlite/ - bionetlite module

  • kernel/ - Neulite kernel (default)

  • flavors/ - Kernel variations

Step 5: Place bionetlite

Place bionetlite.py from the cloned repository in the same directory as your network construction code:

cp neulite/bionetlite/bionetlite.py /path/to/your/network/code/

Directory Structure:

Place bionetlite.py in the same directory level as your network construction script:

your_project/
├── bionetlite.py           # Copy here
├── build_network.py        # Your network construction script
└── config.json             # Configuration file (if needed)

With this arrangement, you can import in build_network.py as follows:

# build_network.py
from bionetlite import NeuliteBuilder as NetworkBuilder

net = NetworkBuilder('my_network')
net.add_nodes(...)
net.build()
net.save_nodes(output_dir='network')

Step 6: Verification

Verify that the placement is correct:

# Move to network construction code directory
cd /path/to/your/network/code/

# Verify that bionetlite.py exists
ls bionetlite.py

# Start Python interpreter to verify
python

>>> from bionetlite import NeuliteBuilder
>>> print("bionetlite is ready to use!")

If no errors are displayed, the setup is successful.

Basic Usage

Converting existing bionet code to bionetlite is very simple. Just change the import statement, and the rest of the code can be used as-is.

Changing the Import

Before (bionet):

from bmtk.builder.networks import NetworkBuilder

After (bionetlite):

from bionetlite import NeuliteBuilder as NetworkBuilder

Network Construction

The rest of the code is exactly the same as bionet:

# Create network
net = NetworkBuilder('v1')

# Add nodes
net.add_nodes(
    N=1,
    pop_name='Scnn1a',
    model_type='biophysical',
    model_template='ctdb:Biophys1.hoc',
    dynamics_params='472363762_fit.json',
    morphology='Scnn1a_473845048_m.swc'
)

# Build network
net.build()

# Save to files
net.save_nodes(output_dir='network')
net.save_edges(output_dir='network')

Confirming Output Files

bionetlite generates files for Neulite in the neulite/ directory:

Directory Structure

neulite/
├── v1_population.csv              # Population information
├── v1_v1_connection.csv           # Connection information
└── data/
    ├── *.swc                       # Processed morphology files
    └── *.csv                       # Ion channel settings

Note

Filenames are generated based on the name specified in NetworkBuilder (in this example, 'v1').

Details of each file:

<network_name>_population.csv - Neuron population information

#n_cell,n_comp,name,swc_file,ion_file

  • #n_cell: Number of cells in the population

  • n_comp: Number of compartments

  • name: Population name

  • swc_file: Path to morphology file

  • ion_file: Path to ion channel configuration file

<src>_<trg>_connection.csv - Synapse connection information

#pre nid,post nid,post cid,weight,tau_decay,tau_rise,erev,delay,e/i

  • #pre nid: Presynaptic neuron ID

  • post nid: Postsynaptic neuron ID

  • post cid: Postsynaptic compartment ID

  • weight: Synaptic weight

  • tau_decay: Decay time constant (ms)

  • tau_rise: Rise time constant (ms)

  • erev: Reversal potential (mV)

  • delay: Transmission delay (ms)

  • e/i: Excitatory (e) or inhibitory (i)

Optional Settings

NeuliteBuilder can accept the following optional parameters:

net = NetworkBuilder(
    'v1',
    convert_morphologies=True,  # Whether to convert SWC files
    convert_ion_channels=True   # Whether to convert ion channel settings
)

convert_morphologies and convert_ion_channels are True by default. In subsequent runs, you can skip conversion by setting these to False.

Details of Dependent Packages

Major Dependent Packages of bionetlite

bionetlite depends on the following Python packages:

bmtk>=1.0.0                    # Brain Modeling Toolkit
neuron>=8.0.0                  # NEURON simulator (required)
numpy>=1.19.0                  # Numerical computing
pandas>=1.2.0                  # Data manipulation
h5py>=3.0.0                    # HDF5 file support
scikit-learn>=0.24.0           # PCA for axon direction estimation (required)
mpi4py>=3.0.0                  # MPI parallel processing (optional)

Required Packages:

pip install bmtk neuron scikit-learn

Optional Packages:

# When using MPI parallel processing
pip install mpi4py

Next Steps

References