Welcome to the PyFibers documentation!

PyFibers

PyFibers paper citation will be available soon.

DOI Contributors Python PyPI PyPI - Downloads

Development Status Code Style Pre-commit Documentation Issues Pull Requests CI

This package implements biophysical models of axons in the NEURON simulation environment using Python. With our package, you can model the responses of axons to electrical stimulation (e.g., find the minimum current amplitude required to activate or block an action potential). You can add your own fiber models and simulations protocols. You can use analytical tools for extracellular potentials, or import from finite element models (FEM) such as COMSOL, ANSYS, or FEniCS.

Feature

Description

Flexible stimulation

Support for custom waveforms and extracellular potential distributions

FEM integration

Easy import of high-resolution potentials from finite element simulations

1D and 3D fibers

Support for both straight and curved fiber geometries

Advanced analysis

Built-in threshold search, conduction velocity measurement, and comprehensive data recording

Extensible

Add your own fiber models and simulation protocols

Simulate recording

Simple tools to calculate single fiber action potentials

Library of built-in fiber models

MRG (Myelinated): MRG-discrete, MRG-interpolation, Peña (Small MRG-interpolation)
Sweeney (Myelinated)
Thio (Unmyelinated): Autonomic, Cutaneous
Sundt (Unmyelinated)
Tigerholm (Unmyelinated)
Rattay (Unmyelinated)
Schild (Unmyelinated): Schild 1994, Schild 1997

Installation

Note that these installation instructions are for users. Developer instructions are available in contributing.md.

It is recommended (But not required) you create a new virtual environment for PyFibers. For example, using Anaconda/Miniconda:

  • conda create -n pyfibers

  • conda activate pyfibers

  1. Install NEURON and add to PATH (https://nrn.readthedocs.io/en/latest/)

    • Make sure your NEURON and Python versions are compatible (https://nrn.readthedocs.io/en/latest/changelog.html)

    • Check your installation by running the following command: python -c "import neuron; neuron.test(); quit()". If successful, test outputs along with “OK” should be printed to the terminal.

  2. Install PyFibers from PyPI and compile the .mod files.

    pip install pyfibers
pyfibers_compile

Some notes for pyfibers_compile:

  • It is normal to see the following message during compilation: NEURON mechanisms not found in <path>. Check the NEURON output that follows for a message that the mechanisms were compiled successfully (e.g., for Windows: nrnmech.dll was built successfully.) In the future while using PyFibers, if you see the NEURON mechanisms not found in <path>. message, this is cause for concern, as this means PyFibers cannot find the compiled mechanisms. Failed compiles will commonly cause the error message Argument not a density mechanism name to appear when trying to create fibers.

  • Careful! Make sure that the correct NEURON installation is in your path, as the first found installation will be used for compilation. The version used for compilation must be the same version used to run PyFibers code.

  • If you receive a message that the pyfibers_compile command is not found, find the executable for this command in the Scripts path of your python directory (e.g. C:\Users\<username>\Anaconda3\envs\pyfibers\Scripts) and run the executable (e.g., pyfibers_compile.exe).

Usage

📖 Documentation: For detailed information on usage, see our documentation.

The basic steps for running a PyFibers simulation are as follows:

Creating a model fiber

Use the build_fiber function to create fiber object. The fiber object consists of NEURON sections with ion channel mechanisms inserted for the fiber model chosen when the object is initialized. Users can add custom fiber models as well as using our provided models (See Custom Fiber Models)

from pyfibers import build_fiber

fiber = build_fiber(
    fiber_model=FiberModel.MRG_DISCRETE,
    diameter=10,  # um
    n_nodes=25,  # um
    temperature=37,  # C
)

Running a Simulation

The fiber object can be run in simulations using our provided ScaledStim class; alternatively, users can create their own custom simulation setup (See Custom Simulations). Once you have a fiber object, you can create a ScaledStim instance, which is a set of instructions for stimulating model fibers.

# Add extracellular potentials to the fiber.
fiber.potentials = potential_values

# Create instance of ScaledStim class
stimulation = ScaledStim(waveform=waveform, dt=time_step, tstop=time_stop)

# run threshold search
amp, _ = stimulation.find_threshold(fiber)
print(f"Threshold for 5.7 micron fiber: {amp} (mA)")

For more examples, see the documentation.

Contributing

If you develop additional functionality that would be generally useful to our users, please open a Pull Request for us to review. For detailed contribution guidelines, see our contributing guide.

Authors and acknowledgment

  • Developed at Duke University by Daniel Marshall, Elie Farrah, and Eric Musselman

  • Please see: <Paper REF Forthcoming>

License

See LICENSE

Getting Started

User Guide

Reference

External Links