Tutorials and Examples

Basic Usage

Tools for Modeling: The Brain Modeling Toolkit and SONATA data format

A full guide for how to build, simulate, and analyze brain networks with (BMTK) using a scaled-down model of the mouse visual primary cortex layer 4 system and a variety of simple and complex stimuli.

tutorials/Mouse_L4.html

BMTK Builder (A Quick Introduction)

A in-depth tutorial for how to use the BMTK Network Builder module to create and save brain networks models across differfent scales and levels-of-resolution.

tutorials/NetworkBuilder_Intro.html

Multi-Population Recurrent Network (with BioNet)

How to use BMTK BioNet to run simulations of networks of biophyscially realistic compartmental cell models.

tutorials/tutorial_04_multi_pop.html

Point-Neuron Network Models (with PointNet)

How to use BMTK PointNet for running simulation of single point-neuron models.

tutorials/tutorial_05_pointnet_modeling.html

Modeling the Visual Field (with FilterNet)

Use BMTK FilterNet to convert stimuli into a series of spikes for analysis and network stimuli.

tutorials/tutorial_07_filter_models.html

Population Level Modeling (with PopNet)

How to use BMTK PopNet for running simulation of populations and population firing rates using DiPDE.

tutorials/tutorial_06_population_modeling.html

More Features

Tools for Generating Cell Placements

• Advanced options and functions for placing cells when building a network using BMTK Network Builder.

• Importing cell locations using Allen Common Coordinate Framework with NNRD files.

tutorials/cell_placement.html

Auditory Stimuli Inputs using FilterNet

How to use FilterNet module to use audio wav files as stimuli for virtual neurons with filters that detect spectral and temporal modulation

tutorials/auditory_filternet.html

Synaptic Plasticity in Pointnet (STP, Facilitation, Depression, STDP, and Others)

• static synapse

• adjusting parameters in PointNet (NEST)

• STDP and STP synpatic models in PointNet (NEST)

tutorials/dynamic_synapses.html

Generating Spike Trains with a Refractory Period

Using built-in Spike Generator process to create spike trains (for analysis and simulation) with Poisson and Gamma based distributions with

• homogeneous firing rates

• Heterogeneous firing-rates

• built-in refractor period

tutorials/dynamic_synapses.html

Replaying Parts of a Simulation

• Use activity of previous recurrent recordings for input to simulation.

• Allows you to capture and separate out network activity not generated by external stimulus.

• Can also select subpopulations of cells and synapses to segment subnetwork and motif activity within a larger network.

tutorials/bionet_disconnected_sims.html

Using Customized and External Cell and Channel Models in BioNet

• Importing NEURON HOC template cell models.

• Overwriting and appending to default cell model parameters and mechanisms.

• Writing custom cell models in Python.

• Importing customized channels and ion mechanisms into existing models.

tutorials/Ch_External_Models.html

Extending PointNet Networks and Simulation with External and Customized Models

• Using Built-in NEST cell models.

• Overridding cell model instantiation.

• Custom cell models with NESTML

tutorials/Ch_NEST_Cusom_Models.html

Extending PointNet Networks and Simulation with External and Customized Models

• Using Built-in NEST cell models.

• Overridding cell model instantiation.

• Custom cell models with NESTML

tutorials/Ch_NEST_Cusom_Models.html

Advanced Methods for Driving your Network with Synaptic Spike-trains

• How to create you’re own pregenerated spike-train files using SONATA, CSV, or NWB files.

• Write your own python function to dynamically generate input spike-trains before or during simulation.

tutorials/Ch_advanced_spikes_input.html

Advanced Stimulus Options

• Advanced Options for setting current clamp stimulus in a simulation.

• Voltage Clamping stimulus.

• Use Allen Cell-Types intracullar experimental stimuli (sweep) for a simulation.

• Extracellular stimulus input

tutorials/Ch_advanced_stim.html

Advanced Stimulus Options

• Recording single and group cell contribution to a extracellular electrode or mesh.

• Setting extracellular resistance.

• Calculating Current Source Density.

tutorials/Ch_extracellular.html

External Resources for Large-Scale Network Modeling

• Running BMTK on the Neuroscience Gateway (NSG).

tutorials/Ch_extracellular.html

BMTK Example Networks

The BMTK github repository also includes a number of examples models that showcase the numerious capabilities and features supported by BMTK, each sub-directory containing a different model including necessary files to build, simulate, and analyize. While most of these are small toy models designed to run quickly on one’s machine, they can provide a good jumping off point for building large more realistic networks, plus more concrete examples of how to use BMTK.

name	description	features
bio_14cells/	A small network of 14 cells - 10 multi-compartment biophysically detailed cells and 4 point integrate-and-fire cells - called V1. Recieves input from two networks of virtual cells (spike-trains) - LGN and TW.	BioNet Builder
bio_1cell/	A single excitatory cell that is stimulated by current clamps or virtual synapses	BioNet Builder IClamp
bio_450cells/	This is a small example network of a 450 cell simulation based on the 45,000 mouse layer 4 network described in Arkhipov et. al. 2018. Of the cells 180 are biophysically detailed multicompartment cells downloaded from the Allen Cell-Types database, the remaining are point integrate-and-fire neuron models. The network is driven by an external network of virtual nodes/spike-trains	BioNet Builder LFP/ECP Membrane (V)
bio_450cells_replay/	This is a small example network of a 450 cell as above, but used to replay module for recreating and isolating network activity of different subsystems.	BioNet Network Replay
bio_advanced_stimuli/	A small toy network with various advanced options for stimulation.	BioNet Extracellular Stimulation IClamp Voltage Clamps Spontaneous Stimuli
bio_all_active_sweep/	An example of using Allen Institute All Active cells models, recreating Allen Cell-Types Sweep experiments.	BioNet model_processing Allen Cell-Types Database
bio_comsol/	An example of advanced extracellular network stimulation using COMSOL physics files.	BioNet Extracellular Stimulation COMSOL
bio_neuropixels/	An example using Neurodata without Borders (NWB) files and DANDI archive for integrating ECEPhys experimental data into a network simulation.	BioNet Builder NWB 2.0 Dandi
bio_nsg/	Includes Files and instructions for running a small network simulation on Neuroscience Gateway (NSG)	BioNet NSG
bio_stp_models/	A network that uses STP type synapses	BioNet Builder STP synapses
point_120cells/	A small network of 120 point-neurons. Uses PointNet and will require NEST to run.	PointNet Builder IClamp Optogenetic inhibition
point_120cells_nestml/	A small network of 120 point-neurons. Neuron and synapse models are specified as NESTML models to demonstrate how to incorporate NESTML models. Uses PointNet and will require NEST and NESTML to run.	PointNet Builder NESTML
point_450cells/	This is a small example network of a 450 point neuron simulation based on the 45,000 mouse layer 4 network described in Arkhipov et. al. 2018. The network is driven by an external network of virtual nodes/spike-trains.	PointNet Builder
point_450glifs/	This is a small example network that uses the Allen Institute generalized leaky integrate-and-fire (glif) cell models. The network is recurrent and receives stimulation from external spike trains (virtual cells).	PointNet Builder Parallel simulations
point_iclamp/	A small example network that uses different types of current clamp (IClamp) stimuli formats.	PointNet IClamp
point_stdp/	PointNet simulation with plastic synapses	PointNet Builder
filter_graitings/	An example for a drifiting grating stimulus. All parameters can be found in the config file	FilterNet Builder Drifting Gratings
filter_looming/	This is an example for creating a looming stimulus from LGN filters. The stimulus is a black circle that expands at 80degs/sec and then repeats. The example uses 85 cells from 3 LGN models.	FilterNet Builder Looming Movie
filter_movie/	FilterNet simulations from arbitary movies	FilterNet Movie
pop_2pops/	Simple Excitatory-Inhibitory 2 population network using PopNet (DIPDE)	PopNet Builder
pop_7pops_converted/	A more complex PopNet example converted from the point_120cells/ example	PopNet Builder