We have already added an exitatory and an inhibitory synaptically activated channel to the /cell/dend compartment, and a spike element to the soma, but haven't yet made any use of them.
Usually, we can treat an axon as a simple delay line for the delivery of spike events that last a single time step. Only if we are interested in understanding the details of axonal propagation would it be necessary to model the axon as a series of linked compartments.
The properies of an axon are split between two types of GENESIS objects. Spiking class elements (a spikegen or randomspike) create the spike events, either when Vm crosses a threshold during an action potential (spikegen), or as a random series of events generated at a specified average rate (randomspike). These send SPIKE messages to a synchannel class element (synchan, hebbsynchan, or facsynchan), which contains fields for the propagation delays and synaptic weighting for each synaptic connection.
In order to understand the use of these synaptically activated channels, you will need to read the documentation for Synaptic Connections, synchan, spikegen, and randomspike. It will also be helpful to look at genesis/Scripts/neurokit/prototypes/synchans.g in order to understand the properties of the channels Ex_channel and Inh_channel in /cell/dend.
Then, try this simple exercise:
Modify the simplecell.g script to add a randomspike element with an average spike rate of 200 spikes per second. Connect it to /cell/dend/Ex_channel, and set the soma current injection to zero. If you would like to play with XODUS graphics some more, plot the Ex_channel conductance Gk on another graph.
If you get stuck, look at tutorial4.g ("the hard way") or tutorial5.g (with readcell). In addition to the random spike input, these scripts illustrate the coupling of a cell's spike output to a synchan, by providing a feedback connection from the cell to itself. The next section of this tutorial provides a more realistic exercise that connects two cells to each other to form a pattern generator circuit.
The goal of this exercise is to create a simple network of two cells that fire in alternate bursts. This will be made from two cells derived from the one created in the simplecell simulation, called /cell1 and /cell2. After you feel that you understand simplecell.g and its included files, and have studied the documentation on the use of the synchan and spikegen objects, copy the cells/simplecell files into a directory of your own. Then, make the changes necessary to create a second cell with no current injection, and plot its Vm on the graph in a different color. Of course, the plot will be a flat line, as it is receiving no stimulus.
Then, use what you have learned about synaptic connections to connect the spike output of cell1 to the excitatory synchan of cell2, and the spike output of cell2 to the inhibitory input of cell1. Use an axonal propagation delay of 0.005 seconds for each connection. Finally, experiment with the synaptic weights for each synapse until you can achieve a pattern of alternate bursts of action potentials. To make it easy to change the weights, you may wish to add dialog boxes for entering weights to the control panel.