This is the readme for the research project under development:

Integration of inhibitory and excitatory synaptic events to neurons
Tom Morse in collaboration with Mike Higley
Ted Carnevale, Gordon Shepherd, Jeremy Chang

Integration

Start with the simple dendrite model.  Move it into a file
ball_and_stick.hoc with the idea that we may make it a template.

Add an inhibitory synapse gabaa.mod that is conductance limited
similar to how the new NMDA receptor is conductance limited.  Add
netstims that trigger these.

Keep the option to coat with spines in the back of my mind while performing
first simulations on a smooth dendrite:

First simple combination of excitation and inhibition:

Have an EPSP generated at a fixed position on a spine attached to the
middle of a shaft.  Measure the peak of the EPSP in 1) the originating
spine head and 2) the adjacent location on the shaft as well as 3)
along the shaft and 4) in a spine that moves along the shaft.  Graph
these peak voltages.  Study with an IPSP from inhibition on 1) a spine
or inhibition on 2) the shaft, each of which, in different locations
along the shaft.

case A: IPSP throughout shaft locations, and timing differences

plots: peak voltages along dendrite as function of inhibition location
(3D height plot?) for particular timings (difference between IPSP and
EPSP events.

peak voltages at excitation location as function of inhibition
location and timing (3D height plot?)

peak voltages as function of timing

case B: IPSP in spine head for spine throughout shaft locations and
timing differences

20140424:
To-do
Add a panel for inhib synapses
close panel for conductance graph
add the mod file that records voltages.

start running as planned.

In process: strange things happen when I change the time step now -
weird steps appear/disappear.  Check the new synapses: Maybe new NMDA
and GABAA synapses need their tables reloaded everytime the timestep
changes.

Still need to implement the storage and output of the max voltages
then ready to run in parallel runs.

20140426

Note: the weight in the NMDA.mod file seems to be ignored in the GUI
at least, and the enc.weight is super threshold around 0.005.  The
inc.weight can also cause spiking from rebound from inhibition.

There are weird effects such as when both e and i weight are 5e-4 and
the times are t=2 and 10 ms, the i seems to cause a sub-threshold
rebound (compare with and without inhibition).

20140522

I need to clean up the code where it is sensibly organized.  The
vestige of ABCD window is scary.  I want the code to look like this

1) a hoc file that loads the simple cell.
2) a gui interface hoc file that can open and close different options
to control the simulation to run interactively
3) a parallel simulator hoc file that can drive the simulation over
a range of parameters.  The output (potentially either peak voltage
levels or peak calcium levels) are stored in a format
for v_e_val_i_val_t_lag.dat or ca_e_val_i_val_t_lag.dat:
secname x_loc value
4) matlab code reads the above output files and graphs them

20140529

For some reason I found that the excitatory synapse did not "work"
unless the time that it starts at is set to 3 ms or later?
I will set the default time in the gui, up from 2.

Also the soma was so large L=diam=75 that it was sucking up all the
EPSP current with virtually no rise in voltage so I lowered it to
L=diam=10


20140722
Studies:
EPSP in the spine head (always)
IPSP on the shaft varied position from 0 to 1.
vary timing from +5 to -5 ms by having the EPSP at t=10
and the IPSP from 5, 6, ..., 14, 15.

driver.hoc will run the program init.hoc while writting a file
synapse_start_times.hoc which contains statements like:


ens.start=9 // (with a 1ms delay)
ins.start=4 to 14 // (with a 1ms delay)
strdef outfile
outfile="ens_9_ins_4.dat" // "columns" are the peak voltages at the soma, adjacent shaft, spine head 

----
