Fast Spiking Basket cells (Tzilivaki et al 2019)

 Download zip file   Auto-launch 
Help downloading and running models
Accession:237595
"Interneurons are critical for the proper functioning of neural circuits. While often morphologically complex, dendritic integration and its role in neuronal output have been ignored for decades, treating interneurons as linear point neurons. Exciting new findings suggest that interneuron dendrites support complex, nonlinear computations: sublinear integration of EPSPs in the cerebellum, coupled to supralinear calcium accumulations and supralinear voltage integration in the hippocampus. These findings challenge the point neuron dogma and call for a new theory of interneuron arithmetic. Using detailed, biophysically constrained models, we predict that dendrites of FS basket cells in both hippocampus and mPFC come in two flavors: supralinear, supporting local sodium spikes within large-volume branches and sublinear, in small-volume branches. Synaptic activation of varying sets of these dendrites leads to somatic firing variability that cannot be explained by the point neuron reduction. Instead, a 2-stage Artificial Neural Network (ANN), with both sub- and supralinear hidden nodes, captures most of the variance. We propose that FS basket cells have substantially expanded computational capabilities sub-served by their non-linear dendrites and act as a 2-layer ANN."
Reference:
1 . Tzilivaki A, Kastellakis G, Poirazi P (2019) Challenging the point neuron dogma: FS basket cells as 2-stage nonlinear integrators Nature Communications 10(1):3664 [PubMed]
Model Information (Click on a link to find other models with that property)
Model Type: Neuron or other electrically excitable cell;
Brain Region(s)/Organism: Hippocampus; Prefrontal cortex (PFC);
Cell Type(s): Hippocampus CA3 interneuron basket GABA cell; Neocortex layer 5 interneuron;
Channel(s):
Gap Junctions:
Receptor(s):
Gene(s):
Transmitter(s):
Simulation Environment: NEURON; MATLAB; Python;
Model Concept(s): Active Dendrites; Detailed Neuronal Models;
Implementer(s): Tzilivaki, Alexandra [alexandra.tzilivaki at charite.de]; Kastellakis, George [gkastel at gmail.com];
Search NeuronDB for information about:  Hippocampus CA3 interneuron basket GABA cell;
Challenging the point neuron dogma: FS interneurons as 2 stage integrators

INFORMATION ABOUT THE SIMULATIONS AND CODES. 
--------------------------------------------------------------------------------
-----------------------------------------------------

Dear Colleagues,

In this .pdf file you can find detailed descriptions on the codes used in this work and information about how to run the simulations. 

Hope you will find them useful! 
Best wishes,

Alexandra Tzilivaki

In case you have any questions do not hesitate to contact me at : alexandra.tzilivaki@charite.de  (Twitter: @ATzilivaki)
For questions regarding the network model please also contact Dr. George Kastellakis at: gkastel@gmail.com 



--------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------


1)	Folder Description



-----------------------
a.	ANN
-----------------------

This folder contains the codes and data used to model the Artificial Neural Network models. 
Related to Figure 6, Supplementary Figure 10 and Table 1 of the paper. For further details, please read the Methods section of the paper.

Requirements for running the codes: 
Python 3.6.8
Pytorch 1.0.1.post2 (pytorch.org)

Contents: 
inn_cell.py	This is the main code that simulates the ANNs and generates the output data. 
plot_cells.py	This code generates the regression analysis results and  the figures
CSVDATA	        This folder contains the data (input and output) used to train the ANNs
data	        The outputs generated by running the inn_cell.py will be saved in this folder. 
run_cells.sh	This bash script runs the  inn_cell.py code and enables the user to choose which of the 4 ANN formalisms (bimodal nonlinear, linear, only supralinear, only sublinear) wish to run and choose the number of epochs and seed 

How to run (open a terminal and type):
> conda activate 
>sh run_cells.py
>python plot_cells.py 




--------------------------
b.	Network Model: 
--------------------------

This folder contains the network simulator that simulates memory engram formation 
in a population consisting of excitatory and inhibitory neurons with independent
dendritic subunits.
Related to Figure 7 and Supplementary Figure 11 of the paper.
 For further details, please read the Methods section of the paper as well as the Kastellakis, G., Silva, A. J., & Poirazi, P. (2016). Linking memories across time via neuronal and dendritic overlaps in model neurons with active dendrites. Cell reports, 17(6), 1491-1504. 
Requirements for running the codes: 
gcc 4.4.7 
python 2.7
GNU Make 3.81


Contents/Directory layout:

data/	              Contains simulation output data. Used to generate figures
src/  	              Contains the implementation of the simulator. Specifically:
src/lamodel.cpp	      Main simulator entry point with command line option parsing
src/constructs.h      Data structure definitions
src/constructs.cpp    Implementation file of simulation dynamics / connectivity and plasticity
src/tests.cpp	      Unit tests
figs/	              Figure-generating python scripts (requires the simulator output data)
figs/engrams.py	      Generates main text figure and the .txt files for the figure data
figs/supl.py	      Generates supplemental figure
run_simulations.sh    Script to run all simulations serially
submit_lamodel.sh     Submission script used to run the simulations in a PBS compatible cluster  (not used by default)

How to run (open a terminal and type):
(To compile the simulator and generate data:)
> make -C src clean all
> sh run_simulations.sh
(To generate figures)
> cd figs
> python  engrams.py
> python  supl.py
(To run unit tests: ) 
> ./tests





--------------------------------------------
c.Multi-compartmental Biophysical Models
--------------------------------------------


This folder contains the codes and data used to model Multi-compartmental biophysical models of the Fast Spiking Basket cells. 
Related to Figures 2, 3 and 4 and Supplementary Figures 1-9 and 12 of the paper. For further details, please read the Methods section of the paper.

Requirements for running the codes:
NEURON 7.4 version
MATLAB 

For more help on downloading and running the NEURON codes (.hoc files), please read here: https://senselab.med.yale.edu/ModelDB/NEURON_DwnldGuide.cshtml

Sub-folders Description:


1) bash_templates
basic_graphics.hoc	(This file creates graphs while running NEURON simulations.)
current_balance_fs.hoc	(This file sets the resting membrane potential to -68 mV.)

2) mechanism
This file contains all the .mod files used to build the multicompartmetal models. For more information about the .mod files please read the methods of the paper. 
*** IMPORTANT*** 
After downloading the Multicompartmental_Biophysical _models file, a successful compilation of the .mod files is needed in order to run the codes in the experiment folder. 

How to compile the .mod files (please open a terminal and type:)
>cd mechanism
>nrnivmodl

3) experiment
Contents:
disperse.hoc	                This files simulates synapses in a randomly disperse pattern in the dendrites (see also in the Methods of the paper: disperse synaptic allocation.)
disperse	                Executable file for running the disperse.hoc code

in_branch.hoc	                This files simulates the grouped synapses in a few randomly selected dendritic branches.  (see also in the Methods of the paper: Grouped synaptic allocation.)
in_branch	                Executable file for running the in_branch.hoc code

Model_validation.hoc	        This files is used to apply multiple current-clamp configurations to the models for validation tests.
model_validation	        Executable file for running the Model_validation.hoc code

IO.hoc	                        This file is used to activate synapses on each of the dendrites respectively.
io	                        Executable file for running the IO.hoc code

gap.hoc	                        This file is used to implement gap junctions in the model cells.
vecstimgap.hoc	                This file is used to implement a firing frequency ~ 30 Hz to the presynaptic cell 
experiment_with_gap.hoc 	This file simulates gap junctions and records dendritic responses.
rungap	                        Executable file for running the experiment_with_gap.hoc code

Vecstim.hoc	                This file simulates synaptic input that lead to ~ 3 Hz (background activity)

Dend_InputResistance.hoc	This code simulate current clamp on the dendrites to calculate input resistance.
dendinputresistance	        Executable file for running the Dend_InputResistance.hoc code

SynCurrents_Validation.hoc	This file simulates multiple voltage clamp configurations for validating the synaptic currents of the model cells.
Syncurrents	                Executable file for running the SynCurrents_Validation.hoc code

PFCtemplate.hoc	                Template with the membrane properties used for the 3 cortical FS BC reconstructions. 
tempSomogyi1.hoc	        Template with the membrane properties used for the Hippocampal Somogyi_1.hoc reconstruction. 
tempSomogyi23.hoc	        Template with the membrane properties used for the Hippocampal Somogyi_2.hoc and Somogyi_3.hoc reconstructions.
tempSomogyi45.hoc	        Template with the membrane properties used for the Hippocampal Somogyi_4.hoc and Somogyi_5.hoc reconstructions.

Model_reconstructions	This folder contains all the morphological FS BCs reconstructions, used in this study.


If you wish to run a particular protocol (e.g the disperse.hoc) please open a terminal and type:
>cd experiment
>./disperse


4)	Figure 2
This folder contains the datasets (in the .mat files) and the script used to generate Figure 2.
5)	Figure 3
This folder contains the datasets (in the .mat files) and the script used to generate Figure 3.
6)	Figure 4
This folder contains the datasets (in the .mat files) and the script used to generate Figure 3.



Loading data, please wait...