Page 1 of 1

### Dominant Tau inTsodyks-Markram synapse model..?

Posted: Mon Sep 10, 2012 5:49 pm
Dear All,

One of the interesting points of Tsodyks-Markram model is the possibility to include facilitation,

This way three states are defined

x=> fraction of available synaptic resources but not active
y=> fraction or active synaptic resources
z=> fraction of inactive resources.

Such that
y converts to z with time constant t1, and z converts back to x with time constant trec

I want to study the non-facilitated problem and for that

I assume new y0 = y0 + x instead of new y0 = y0 + x · u with u the facilitation term. this because It results a very good and more realistic approach to study of the GABAA dynamics. than just one mono-exponential.

However, since this is a multi-state approach the total time of activation should be comparable with the tau in the mono-exponential case for the limit case, but I can not figure out how to set conditions for that limit case:

Should I assume that trec dominates the dynamics of the channel? or is it somehow a combination between t1 and trec?

Thanks a lot for any feedback!!

Best,

OJAG

### Re: Dominant Tau inTsodyks-Markram synapse model..?

Posted: Wed Sep 12, 2012 4:46 pm
OJAG wrote:I want to study the non-facilitated problem and for that

I assume new y0 = y0 + x instead of new y0 = y0 + x · u with u the facilitation term.
If you are using my implementation of that mechanism, all that is necessary is to set tau_facil to 0. Then the constant parameter U controls the fraction of x ("available transmitter") that is "released." If you want all of it to be released, just set U to 1.
Should I assume that trec dominates the dynamics of the channel? or is it somehow a combination between t1 and trec?
The synaptic conductance decays with a single time constant: tau_1. tau_rec merely governs the rate at which the amount of "available transmitter" is restored.