Insert Point Process in NetworkBuilder

Moderator: wwlytton

Post Reply
Johannes

Insert Point Process in NetworkBuilder

Post by Johannes » Thu Jun 02, 2005 2:46 pm

Hi,

I have 2 questions:

1) I created a very simple network (2cells connected via one ExpSyn) with the network builder. Now I need to insert a point process (VClamp) at the presynaptic neuron, yet I can't find the created network cells in the point processes menu...

2) Alternatively, I wrote the same network as a .hoc file and then started up the GUI. Now I can't bring up a menu to change the synaptic weight!?
Help, please!

Thanks, Johannes

ted
Site Admin
Posts: 5598
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: Insert Point Process in NetworkBuilder

Post by ted » Thu Jun 02, 2005 3:50 pm

Johannes wrote: 1) I created a very simple network (2cells connected via one ExpSyn) with the network builder. Now I need to insert a point process (VClamp) at the presynaptic neuron, yet I can't find the created network cells in the point processes menu...
Several issues here.

1. Point processes menu? What point processes menu? How would a menu
of point processes help you attach a point process to a biophysical model cell?

2. Are you sure you want a VClamp? Only use VClamp if one or more of the
following is true:
--you need a two electrode voltage clamp
--you need to be able to control the gain and bandwidth of the clamp
Otherwise use SEClamp (Single Electrode Clamp). Computationally more
efficient, fewer details to have to manage.

3. When you're dealing with a net created by the Network Builder, the
GUI's Shape plots (including the Shape plots of the PointProcessManager and
PointGroupManager) will show the "NetStim host section" acell_home_ in
addition to whatever biophysical model cells are in your network. Furthermore,
a new PPM will come up with acell_home_ as the selected section. This can
be puzzling, especially if your biophysical model cell is small compared to
acell_home_ (which has L 100 and diam 500). Often, the biophysical model
cells are shown as very tiny dots. If this happens, it may be necessary to
zoom in to be able to even see the biophysical model cell(s), let alone place
a point process on one of their sections.

4. If you prefer to manage point processes by writing hoc code, you need
to discover the real names of the cells in your network. The Network Builder
and its related tools display short aliases, because the real (hoc-callable) names
are often very long and would take up too much space in these GUI tools.
To discover the real names of things, click on the Network Builder's Net Variables
button and select Show Map. In a toy net that I'm looking at, I see that the
biophysical model cell is called pyr in the Network Builder, but its real name
in hoc is pyr_Cell[0]. This is a specific instance of the class pyr_Cell, so
sections that belong to pyr_Cell[0] are called pyr_Cell[0].soma, pyr_Cell[0].dend[0]
etc..
2) Alternatively, I wrote the same network as a .hoc file and then started up the GUI. Now I can't bring up a menu to change the synaptic weight!?
If you aren't using the Network Builder to create and manage your net, you aren't
going to be able to use its superkludgy weights tool to adjust synaptic weights.
You'll have to adjust weights from hoc. If that's too tedious, you can always
create your own xpanel and put a few xvalue statements in it that would give
you quick and easy control over weights. You can read about xpanel and xvalue
in the Programmer's Reference.

Post Reply