A two networks model of connectivity-dependent oscillatory activity (Avella OJ et al. 2014)

 Download zip file   Auto-launch 
Help downloading and running models
Activity in a cortical network may express a single oscillation frequency, alternate between two or more distinct frequencies, or continually express multiple frequencies. In addition, oscillation amplitude may fluctuate over time. Interactions between oscillatory networks may contribute, but their effects are poorly known. Here, we created a two model networks, one generating on its own a relatively slow frequency (slow network) and one generating a fast frequency (fast network). We chose the slow or the fast network as source network projecting feed-forward connections to the other, or target network, and systematically investigated how type and strength of inter-network connections affected target network activity. Our results strongly depended on three factors: the type of the relevant (main) connection, its strength and the amount of source synapses. For high inter-network connection strengths, we found that the source network could completely impose its rhythm on the target network. Interestingly, the slow network was more effective at imposing its rhythm on the fast network than the other way around. The strongest entrainment occurred when excitatory cells of the slow network projected to excitatory or inhibitory cells of the fast network. Just as observed in rat activity at the prefrontal cortex satisfies the behavior described above, such that together, our results suggest that input from other oscillating networks may markedly alter a network’s frequency spectrum and may partly be responsible for the rich repertoire of temporal oscillation patterns observed in the brain.
1 . Avella Gonzalez OJ, van Aerde KI, Mansvelder HD, van Pelt J, van Ooyen A (2014) Inter-network interactions: impact of connections between oscillatory neuronal networks on oscillation frequency and pattern. PLoS One 9:e100899 [PubMed]
Citations  Citation Browser
Model Information (Click on a link to find other models with that property)
Model Type: Realistic Network;
Brain Region(s)/Organism:
Cell Type(s):
Gap Junctions:
Receptor(s): GabaA; AMPA;
Simulation Environment: NEURON;
Model Concept(s): Oscillations; Brain Rhythms;
Implementer(s): Avella G. Oscar Javier [oscarjavella at gmail dot com];
Search NeuronDB for information about:  GabaA; AMPA;
Author: Oscar Javier Avella Gonzalez (oscarjavella at gmail.com)

Paper: Avella Gonzalez, O. J., Van Aerde, K. I., Mansvelder, H. D.,
Van Pelt, J., and Van Ooyen, A. (2014). Inter-network interactions:
impact of connections between oscillatory neuronal networks on
oscillation frequency and pattern. PloS ONE 9(7):e100899

The model was fully implemented in NEURON, with output files and
analysis tools in Matlab.
!!! Contains the MultipleRunControl, originally designed by R.A.J van
Elburg The software is released under the GNU GPL version 3:

Purpose of this work:

The model was created to study the possible effects on the oscillatory
dynamics of a neuronal network (target network) when receiving
feed-forward synaptic input from a second oscillatory network (source
network). We simulated two non-harmonic oscillating networks, both
consisting of a population of excitatory (E) cells and a population of
inhibitory (I) cells. Both types of neurons are designed as single
compartments, each characterized by the type of synapse it projects
onto the postsynaptic cell, while membrane kinetics of both cell types
are the same. Thus, E cells project AMPA synapses, while I cells
project GABA_A synapses. The cells belonging to a certain network are
connected with specific probabilities, creating in each network a
pyramidal interneuron gamma-like oscillation (PING), whose oscillation
frequency depends on the decay time of the inhibitory channels.

The conductance factor C_f used to modify the strength of the
conductance of the main connections between the two networks was
varied as follows C_f=0, 0.01, 0.05, 0.5, 0.8, 1.5, 3.0, 7.0, 10.0,
15.0, 20.0.

For further details concerning the simulation and specific parameters,
please check the Methods section in the paper above.

Three types of outputs emerged from the model activity depending on
the strength of the "main" internetwork synapse, its type and number:

a) Locking of the activity of the target network to that of the source
b) Transitions between episodes of high and low amplitude oscillations
c) Interspersed activity of two different rhythms.

Running the simulation:

The model was originally ran on NEURON 7.0 using the conventional
commands line scheme, edited in pspad, but any other text editor also
works.  To set up the simulation example, go to /main directory and
load the file


Once the windows are open, press the button "single run" in the
MultipleRuncontrol panel. The simulation will start, running for
40000ms. When the simulation stops, expand the window "Pyram Cells
Population # 1" in the horizontal axis between 16 and 18s.

Producing the example figure:

Open the /main menu and load, double clicking on the file
"runExampleFigs.m" Once Matlab has opened, press the button "run".


The figure corresponds to the data in file
.\main\output_matlab\g_eE_conn_ei2E_exp4_Ready.m, with current files
set in the NEURON model.

The top panels represent the Fourier transform of the slow network
(source) and corresponding wavelet transform during the time of
simulation; and the bottom panels are the same for the fast network,
in this case the target network.  You can compare the
Example_figure.tif with the d. and e. panels in Fig. 9 of the paper.
Observe the resulting interspersed activity of the target network as a
direct consequence of the feed-forward connections projecting from the
excitatory (e) and inhibitory (i) cells of the slow network to the
excitatory (E) cells of the fast network, having e to E connection as
the relevant internetwork connection, where "e" and "i" stand for
excitatory/inhibitory cells of the slow network and "E" and "I" for
those of the fast network and Cf=0.8.

Manipulating the files:

You can change the Cf parameter by editing the
/main/functions_net_multipop.hoc file and determine which connections
between the networks must be on during the simulation.

-check lines 43 to 51   
geI_factor=1//multiplicative factors of the 
giI_factor=0//conductances used to connect 
geE_factor=0.8//intersecting cells of both networks


and proceed as follows:

For the main connection, vary the g_(SourceTarget)_factor Cf with
Source/Target=i,e,I,E , i.e. each of the 64 possible combinations
using the values Cf=(0.01, 0.05, 0.5, 0.8, 1.5, 3.0, 7.0, 10.0, 15.0,

For complementary connection, set the respective
"g_(source_target)_factor" to 0 if not included or 1 if included.

To change the name of the output .m file, open the file
\main\sessions\DrivePower_run_multi.ses and follow instructions of
lines 40-42.


To reproduce the dot figures of the paper, run the file
/Matlab_support/spikeTrainAnalysis/analisis1.m using the structure
filename g_(MainConnection)conn_(source)2(target)exp(number) ej:
g_eE_conn_ei2E_exp4, where number stands for the associated index in
the list below for the corresponding C_f factor= (0.01, 0.05, 0.5,
0.8, 1.5, 3.0, 7.0, 10.0, 15.0, 20.0).