The code in this notebook implements the down-scaled version of spiking network model of macaque visual cortex developed at the Institute of Neuroscience and Medicine (INM-6), Research Center Jülich. The full-scale model has been documented in the following publications:
1. Schmidt M, Bakker R, Hilgetag CC, Diesmann M & van Albada SJ
Multi-scale account of the network structure of macaque visual cortex
Brain Structure and Function (2018), 223: 1409 [https://doi.org/10.1007/s00429-017-1554-4](https://doi.org/10.1007/s00429-017-1554-4)
2. Schuecker J, Schmidt M, van Albada SJ, Diesmann M & Helias M (2017)
Fundamental Activity Constraints Lead to Specific Interpretations of the Connectome.
scale_down_to is the down-scaling factor which defines the the ratio of the full scale multi-area model being down-scaled to a model with fewer neurons and indegrees so as to be simulated on machines with lower computational ability and the simulation results can be obtained within relative shorter period of time. <br> Its deafualt value if `1.` meaning full scale simulation. <br> In the pre-set downscale version, it's set as `0.005`, where the numer of neurons and indegrees are both scaled down to 0.5% of its full scale amount, where the model can usually be simulated on a local machine. <br>**Warning**: This will not yield reasonable dynamical results from the network and is only meant to demonstrate the simulation workflow <br>
2.`cc_weights_factor`<br>
This scaling factor controls the cortico-cortical synaptic strength. <br> By default it's set as `1.0`, where the inter-area synaptic strength is the same as the intra-areal. <br>**Important**: This factor changes the network activity from ground state to metastable state. <br>
3.`areas_simulated`<br>
This parameter specifies the cortical areas included in the simulation process. Its default value is `complete_area_list` meaning all the areas in the complete_area_list will be actually simulated. <br>
The value assigned to simulation_areas can be any sublist of the compete_area_list specifying areas a user want to include in his/her simulation. <br>
4.`replace_non_simulated_areas`<br>
The paramter replace_non_simulated_areas defines how non-simulated areas will be replaced. <br> It's set as `None` by default when the parameter areas_simulated is set as full_area_list where all areas will be simulated so that no areas need to be replaced. <br> Other options are: `'hom_poisson_stat'`, `'het_poisson_stat'`, and `'het_current_nonstat'`. `'hom_poisson_stat'` is a manually set parameter which can be tuned. When it's set as 'het_poisson_stat' or 'het_current_nonstat', the data to replace the cortico-cortical input is loaded from 'replace_cc_input_source' which is the firing rates of our full scale simulation results. The differenc between 'het_poisson_stat' and 'het_current_nonstat' is that 'het_poisson_stat' is the mean of the time-series firing rate so that it's static, yet 'het_current_nonstat' is time-varying specific current, which is varying by time.
%% Cell type:code id:60265d52 tags:
``` python
# Downscaling factor
# Value range/options: (0, 1.]
# Value assgined: 0.005
scale_down_to=0.005# Change it to 1. for running the fullscale network
# Scaling factor for cortico-cortical connections (chi)
# Value range/options: [1., 2.5]
# Value assgined: 1.0
cc_weights_factor=1.0
# Cortical areas included in the simulation
# Value range/options: any sublist of complete_ares_list
We try our best not to confuse users with too many parameters. However, if you want to change more parameters and explore the model, you can do so by passing a dictionary to the `default_params` argument of the `MultiAreaModel` class.
%% Cell type:code id:6e4bed8d tags:
``` python
# Connection parameters
conn_params={
'replace_non_simulated_areas':'het_poisson_stat',# Whether to replace non-simulated areas by Poisson sources with the same global rate, by default: None
'g':-11.,# It sets the relative inhibitory synaptic strength, by default: -16.
'K_stable':'K_stable.npy',# Whether to apply the stabilization method of Schuecker, Schmidt et al. (2017), by default: None
'fac_nu_ext_TH':1.2,# Increase the external input to 2/3E and 5E in area TH
'fac_nu_ext_5E':1.125,# Increase the external Poisson indegree onto 5E
'fac_nu_ext_6E':1.41666667,# Increase the external Poisson indegree onto 6E
'av_indegree_V1':3950.# Adjust the average indegree in V1 based on monkey data
}
# Input parameters
input_params={
'rate_ext':10.# Rate of the Poissonian spike generator (in spikes/s)
}
# Neuron parameters
neuron_params={
'V0_mean':-150.,# Mean for the distribution of initial membrane potentials, by default: -100.
'V0_sd':50.}# Standard deviation for the distribution of initial membrane potentials, by default: 50.
# Network parameters
network_params={
'N_scaling':scale_down_to,# Scaling of population sizes, by default: 1.
'K_scaling':scale_down_to,# Scaling of indegrees, by default: 1.
'fullscale_rates':'tests/fullscale_rates.json',# Absolute path to the file holding full-scale rates for scaling synaptic weights, by default: None
### 2.2. Predict firing rates from theory <a class="anchor" id="section_2_2"></a>
%% Cell type:code id:6a7ddf0e tags:
``` python
p,r=M.theory.integrate_siegert()
print("Mean-field theory predicts an average "
"firing rate of {0:.3f} spikes/s across all populations.".format(np.mean(r[:,-1])))
```
%% Cell type:markdown id:2062ddf3 tags:
### 2.3. Extract and visualize interareal connectivity <a class="anchor" id="section_2_3"></a>
%% Cell type:markdown id:8a7c09e0 tags:
The connectivity and neuron numbers are stored in the attributes of the model class. Neuron numbers are stored in `M.N` as a dictionary (and in `M.N_vec` as an array), indegrees in `M.K` as a dictionary (and in `M.K_matrix` as an array). Number of synapses can also be access via `M.synapses` (and in `M.syn_matrix` as an array). <br>
