NetCon detection threshold
Posted: Mon Jul 18, 2011 10:12 am
Hello,
I'm working on a model of dendro-dendritic interaction with the following code.
From the cell template which acts as a source:
Create and register source cell:
Connecting the source compartment to some other cell's target compartment:
So far so good, but now I have experienced the following weird behavior in a setup where the membrane potential in the source cell fires two spikes. Depending on the NetCons threshold "thresh_dd" the following things happen:
If thresh_dd >= -30, no "spike" or threshold crossing signal is transmitted to the target cell.
If thresh_dd = -40, both of the two "spikes" are detected and a response signal can be seen on the target cell.
If thresh_dd = -35, only the second time the membrane potential in the source compartment crosses the -35 mV is registered and sent to the target cell.
The cells I'm using are both real multi-compartment cells with some HH-type ion channels taken from existing models from ModelDB.
In a serial version this behavior does not appear.
Any ideas what's wrong?
Many thanks in advance!
Best regards,
Bernhard
I'm working on a model of dendro-dendritic interaction with the following code.
From the cell template which acts as a source:
Code: Select all
// this happens in the dendritic compartment:
// append a NetCon object to a public List, with an empty object 'nil' as target, and thresh_dd as threshold for signal detection, zero delay and an arbitrary weight
thresh_dd = -30
src_netcons_glom.append(new NetCon(&v(0.5), nil, thresh_dd, 0, 0.5))
Code: Select all
if (pnm.gid_exists(gid)){ // if this gid has been assigned to the current machine
cell = new Mit(rnd_int.repick()) // create the cell
pnm.register_cell(gid, cell) // register the cell with the gid
pnm.set_gid2node(gid + n_cells, pc.id()) // gid+n_cells is the gid of the NetCon above
pc.cell(gid + n_cells, cell.src_netcons_glom.o(0)) // mark the NetCon from above as output source
}
Code: Select all
td_gid = src_gid + n_cells// the gid of the threshold detecting NetCon object in the source cell's compartment, src_gid is the gid of the source cell
tgt_gid = pg_offset + j // target cell's gid
synapse_id = 0 // synapse id at the source cell in the target compartment
delay = 10 // doesn't matter if 0 ms or anything else
pnm.nc_append(td_gid, tgt_gid, synapse_id, w_mit_pg, delay)
If thresh_dd >= -30, no "spike" or threshold crossing signal is transmitted to the target cell.
If thresh_dd = -40, both of the two "spikes" are detected and a response signal can be seen on the target cell.
If thresh_dd = -35, only the second time the membrane potential in the source compartment crosses the -35 mV is registered and sent to the target cell.
The cells I'm using are both real multi-compartment cells with some HH-type ion channels taken from existing models from ModelDB.
In a serial version this behavior does not appear.
Any ideas what's wrong?
Many thanks in advance!
Best regards,
Bernhard