simulating a spiking neural network

[henrymiller....@gmail.com]

Hi,

I am trying to reproduce the figures in the attached Paper 'Self-adaptive-spike-time-dependednt-pdf, in particular Figure 4c and 4d. (The raster plots and the distribution of initial and final conductances) using 
my try-example in this case, the attached png. Is there a way I can do this in Brian (code)? My problem is understanding how the input spikes are fed into the poisson neurons. Thanks!

Regards,
Henry

[Marcel Stimberg]

Hi Henry,

I think your schematic is confusing things a bit: if the neurons 1–100 are Poisson neurons that spike with 14 Hz, then they do not get any input. Their effect on the post-synaptic neuron could be a change in the post-synaptic membrane potential that looks like the shapes on the left, even though it would be a bit unusual. I don't know the paper and I only had a cursory look, but I have to say it does not describe their model very clearly... If you want to model rectangular EPSPs, you need to use multiple synaptic pathways with different delays, have a look at https://groups.google.com/d/msg/briansupport/WDNyQoVBmX0/SYDUeLz1BAAJ for an example.

Hope that helps, best

  Marcel