Functional Simulation with Point Neurons

This Use Case can be used to test the functionality of the cerebellar microcircuit, in the NEST environment. The Use Case can be found in Online Use Cases/Brain area circuit in silico experiments/Cerebellum/Functional simulation with point neurons.

../../../../_images/schema.png ../../../../_images/functional_flow_chart.png

Approach:

A network is built in pyNEST, following the specificities generated by “Cell positioning” and “Connectome”. The network can then be run (locally and, in the future, also using HPC by UNICORE) imposing input stimulation patterns on specific neurons at the input stage. Each neuron is created as a spiking neuron using IAF dynamics with conductance-based synapses, with parameters that are specific for each neuron type. Pairwise connections are created and tuned in a simplified way (weight positive/negative and delay). An input pattern is defined on glomeruli; the simulation runs, and the spiking activity of all neurons (neuron ID and spike times) is recorded and stored. The network spiking dynamics can be investigated by estimating population frequency-related parameters.

Inputs:

  • a single hdf5 file (saved in /storage) containing position and connection data and some useful parameters from scaffold_params.py (in /storage)

The user can define some features of the, “stimulation input spike patterns”: stimulation area radius (glomeruli within a central area with a tunable radius in µm), start and stop time instants of stimulation (ms), total stimulation duration (ms) and spike frequency (Hz).

Output:

  • Files: .gdf files containing spike times of each network neuron along the simulation duration. The user can also choose to save .dat files containing voltage traces
  • Monitoring: 3D scatter plot of spiking activity within the network; raster plot of spikes for each neuron type (selectable by the user) and PSTH

Additional information:

  • The whole Use Case should take about 12-15 minutes for a volume base of 200 x 200 µm.
  • BBP tool used in the notebook: pyNEST

Details for each neuron model in the pyNEST network Main ref from neuroelectro.org

../../../../_images/table.png

EXAMPLE

A volume of:

  • x = 200 µm, z = 200 µm (→ DCN 100 x 100 µm)
  • y = 930 µm (600+150+30+150 µm), i.e. thickness DCN + GRL+ PCL + ML

THUS with 24157# NUMBER of PLACED NEURONS and 634165# of GENERATED PAIRWISE CONNECTIONS, the PSTH for GOLGI CELLS SPIKES is:

../../../../_images/output.png