oxynet simulates the spiking activity of the magnocellular oxytocin network system, based on the model presented in "Emergent Synchronous Bursting of Oxytocin Neuronal Network" by Rossoni et al. (2008).
For oxynet:
- A C++17 compliant compiler
For plotter.py:
- Python 3 (including numpy, pandas and matplotlib)
In a terminal,
$ git clone https://github.com/rosencode/oxynet.git
$ cd oxynet
$ make install./oxynet [-i parameters_file] [-o spiketimes_file] [-t connections_file] [-g]
Launching oxynet without arguments runs a simulation with the default set of parameters as reported in [1], displaying synchronized bursting.
To change the model parameters, select option -i with the parameter file. The latter should be formatted as a list of (name, value) pairs, as shown below. By default, all the parameters used in a simulation are saved on file parameters.csv.
The network connections are generated randomly upon execution and are saved on file connections.csv (select option -t to specify a different filename).
The seed used to initialize the random number generator is saved with the other simulation parameters, thus allowing to run different simulations with the same network structure.
The connection file is formatted as follows:
- On the first row is the triple (N_c, N_d, N_b) where N_c is the number of neurons, N_d is the number of dendrites on each neuron, and N_b is the number of bundles in the network;
- On each following is a comma-separated list of indices specifying the bundles wherein the i-th cell has its dendrites.
The spiking activity in the network is recorded by default on file spike_times.csv as of a list of (time, neuron_index) pairs preceded by a header row.
Select option -o to redirect the output to a different file.
When the option -g is selected, oxynet initializes and saves the network connections but the execution is stopped before generating the temporal dynamics. This can be useful for studying different network structures.
The input file should be formatted as a list of (name, value) pairs. Parameter names should be exactly as follows:
| Name | Value (default) | Description |
|---|---|---|
| Ncells | 48 | Number of neurons |
| Nbun | 12 | Number of bundles |
| Ndend | 2 | Number of dendrites per cell |
| vrest | -62 | Resting potential (mV) |
| vthre0 | -50 | Spike threshold baseline (mV) |
| tau_m | 10.8 | Membrane time constant (ms) |
| I_su | 0.0005 | Suckling input |
| a_E | 0.0645 | Excitatory PSP amplitude for v = v_E (mV) |
| a_I | 0.22 | Inhibitory PSP amplitude for v = v_E (mV) |
| re | 80 | Excitatory synaptic input rate (Hz) |
| ri | 80 | Inhibitory synaptic input rate (Hz) |
| ahp_max | 40 | AHP, maximum amplitude (mV) |
| ahp_tau | 2000 | AHP, time constant (ms) |
| cann_thre | 0.03 | EC threshold for synaptic attenuation |
| ec_step | 0.0025 | EC increase per unit oxytocin release |
| f_max | 0.6 | Maximum fractional attenuation of synaptic input rate |
| tau_rel | 50 | Maximum interspike interval for release (ms) |
| half_ahp | 45 | AHP, half-activation constant |
| hap_max | 40 | HAP, maximum amplitude (mV) |
| hap_tau | 12.5 | HAP, time constant (ms) |
| T_OT_max | 25 | Maximum OT-induced depolarization (mV) |
| tau_g_ot | 1000 | Time decay of OT-induced depolarization (ms) |
| k_r | 0.045 | Maximum fraction of dendritic stores released per spike |
| ot_to_dep | 0.5 | Depolarization per unit oxytocin release (mV) |
| rel_delay | 5 | Time delay for OT release (ms) |
| tau_cann | 6000 | Time constant for EC decay (ms) |
| tau_r | 400000 | Time constant for priming (ms) |
| t_end | 300000 | Simulation time (ms) |
| random_seed | 1 | Seed for random number generation. Set 0 to randomize at each run |
The file may include as few parameters as required. Note that the current version of oxynet will only run simulations with Ndend = 2, but will still generate network connections with any value Ndend>1.
The Python script plotter.py allows a basic visualization of simulation results. Simply run
python3 plotter.py [spike_datafile]
passing the name of the spike data file to show the firing rates of three sample cells in the network and a raster plot of the spiking activity.
01 March 2021
02 April 2021
Enrico Rossoni ([email protected])
[1] Rossoni et al. (2008) Emergent Synchronous Bursting of Oxytocin Neuronal Network. PLoS Comput Biol 4(7): e1000123. doi:10.1371/journal.pcbi.1000123.