NEURON Files

NEURON simulations

All NEURON simulations are handled using the PyFibers Python package [Marshall et al., 2025] (PyPi, GitHub).

Create fiber model

Based on the “fiber_model” (associated by the “fiber/modes” type parameter in <sim_index>.json (Sim Parameters)), run_controls.py builds an instance of the fiber model using PyFibers. For all fiber types, the segments created and connected in NEURON have lengths that correspond to the coordinates of the input potentials.

Extracellular stimulus

To simulate response of individual fibers to electrical stimulation, we use NEURON’s extracellular mechanisms in the PyFibers package to apply the electric potential from COMSOL at each segment of the cable model as a time-varying signal. run_controls.py loads in the stimulation waveform from a n_sim’s data/inputs/. The saved stimulation waveform is unscaled, meaning the maximum current magnitude at any timestep is +/-1. Analogously, run_control.py reads in the potentials for the fiber being simulated from data/inputs/. ASCENT can run simulation to test for activation threshold, block threshold, or test responses to specific amplitudes (See Simulation Protocols).

Intracellular stimulus

For simulations of block threshold, an intracellular test pulse is delivered via a synapse at one end of the fiber to test if the cuff electrode (i.e., placed between the synapse and the site of detecting action potentials) is blocking action potentials (Simulation Protocols). The intracellular stimulation parameters are defined in Sim and are defined as parameters in NEURON within the run_controls.py file. The parameters in Sim control the pulse delay, pulse width, pulse repetition frequency, pulse amplitude, and node/section index of the intracellular stimulus (Sim Parameters). For simulating activation thresholds, the intracellular stimulation amplitude should be set to zero.

Recording

The NEURON simulation code contains functionality ready to record and save to file the values of state variables at discrete spatial locations for all times and/or at discrete times for all spatial locations (i.e., nodes of Ranvier for myelinated fibers or sections for unmyelinated fibers) for applied extracellular potential, intracellular stimulation amplitude, transmembrane potential, and gating parameters using the PyFibers package. The recording tools are particularly useful for generating data to troubleshoot and visualize simulations.

Save outputs to file

At the end of the NEURON simulation, the program saves state variables as indicated in "saving" in sim.json, CPU time, and threshold values (If running a threshold search) or fiber action potential counts (if running a finite amplitudes protocol). Output files are saved to the data/outputs/ directory within its n_sim folder.

Simulation hierarchy & run_controls.py

The hierarchy of the NEURON simulations can be described as follows:

For each run, the submit.py file is responsible for creating each of the jobs that need to be executed.
For each job, the submit.py file generates and runs shell scripts that handle the submission of the respective jobs.
Finally, the shell scripts call the run_controls.py script, which performs the actual simulation tasks.

The run_controls.py file controls the parameters and simulation protocol for modeling fiber response to electrical stimulation in NEURON and is automatically populated based on parameters in Model and Sim. Parameters defined in <sim_index>.py and <model_index>.json span the categories of: environment (i.e., temperature from Model), simulation time (i.e., time step, duration of simulation from Sim), fiber parameters (i.e., flags for fiber geometry and channels, number of fiber nodes from Model, Sim, and config/system/fiber_z.json), intracellular stimulation (i.e., delay from start of simulation, amplitude, pulse duration, pulse repetition frequency from Sim), extracellular stimulation (i.e., path to waveform file in n_sim/ folder which is always data/inputs/waveform.dat), flags to define the model parameters that should be recorded (i.e., Vm(t), Gating(t), Vm(x), Gating(x) from Sim), the locations at which to record the parameters (nodes of Ranvier for myelinated axons from Sim), and parameters for the bisection search for thresholds (i.e., activation or block protocol, initial upper and lower bounds on the stimulation amplitude for the bisection search, and threshold resolution for the bisection search from Sim). The run_controls.py file takes inputs of the Sim directory, n_sim/ directory, and an exception configuration. The file automatically loads the fiber_z.json configuration file which contains all associated flags, parameters, and rules for defining a fiber’s geometry and channel mechanisms in NEURON, all of which is handled in the PyFibers package.