Stomatogastric ganglion Central Pattern Generator:
 generates pyloric rhythm of the lobster

Network model translated from:
Prinz, Bucher, Marder, Nature Neuroscience, 2004;
STG neuron models translated from:
Prinz, Billimoria, Marder, J.Neurophys., 2003.

Translated into MOOSE by Aditya Gilra, Bangalore, 2013, revised 2014.

On a terminal, cd to the lobster_pyloric path and then execute:
python STG_net.py

You can switch between graded (default) versus spiking synapses by setting the graded_syn variable inside STG_net.py.

The model is fully specified in NeuroML 1.8.1 except the graded synapses which are in pyMOOSE as NeuroML 1.8.1 doesn't support graded synapses.

You can change the model by setting filename in STG_net.py to one of:
Generated.net.xml (default)
Generated.net.reverse.xml (reverse pyloric rhythm)
Generated.net.halfcenter.xml (2 neurons half-center oscillator)
Generated.net.3syn.xml (3 synapses version of pyloric rhythm)

Currently the default network model Generated.net.xml file with default graded synapses (as in the lobster and in the Prinz et al 2004 paper) gives the correct pyloric rhythm.
With default network with spiking synapses (graded_syn=False, un-lobster-like), one of the cells doesn't spike.
For the reversed network: Generated.net.reverse.xml, both the graded and the spiking synapses give the reverse sequence of spiking.

However, one of the synaptic parameters (LP to PY) had to be changed from the original Fig 3e of the Prinz et al 2004 paper.
I think basal Ca and resting potential (to confirm, it's been a while since I coded this) are not specified in the original paper, so that could be the reason of the discrepancy, or could be a bug in my implementation.

