Re: [DS] Matrix for strength of the connection values?

[Salva Ardid]

In the AMPA mechanism, try replacing:

-I1_E1_iAMPA_s(1,:)*(p.I1_E1_iAMPA_gAMPA*I1_E1_iAMPA_netcon).*

by

-I1_E1_iAMPA_s(1,:).*(p.I1_E1_iAMPA_gAMPA*I1_E1_iAMPA_netcon).*

Note .*() instead of *() above

Assuming that dimensions are correct, I think it should work...

El dijous, 2 d’abril de 2020, a les 15:25:11 EDT, Dylan Layfield va escriure:

I have an E-I network for each im trying to set a matrix for the strength
of the connections by changing either the 'gGABAa' value or the 'gAMPA'
value to matrix instead of just a single value

But when I try to put in a matrix of values for the 'gAMPA'

strConEI = [0.1*ones(1,33) 0.2*ones(1,33) 0.1*ones(1,34)]

in the code:


%Setting Equstions for the Model
eqns={
'dv/dt=Iapp+@current+noise*randn(1,N_pop);'
'monitor MECHANISM.functions, monitor v.spikes'
};

%timer
tic;

%create DynaSim specificaiton structure
s=[];
%settings for E cells
s.populations(1).name='E1';
s.populations(1).size=nE;
s.populations(1).equations=eqns;
s.populations(1).mechanism_list={'iNa','iK'};
s.populations(1).parameters={'Iapp',IappVals,'gNa',120,'gK',36,'noise',40};
%settings for I cells
s.populations(2).name='I1';
s.populations(2).size=nI;
s.populations(2).equations=eqns;
s.populations(2).mechanism_list={'iNa','iK'};
s.populations(2).parameters={'Iapp',0,'gNa',120,'gK',36,'noise',40};
%settings for I to E connections
s.connections(1).direction='I1->E1';
s.connections(1).mechanism_list={'iGABAa'};
s.connections(1).parameters={'tauD',5.5,'gGABAa',0.1,'netcon',netConIE};
%settings for E to I connections
s.connections(2).direction='E1->I1';
s.connections(2).mechanism_list={'iAMPA'};
s.connections(2).parameters={'tauD',2,'gAMPA',strConEI,'netcon',netConEI};
data=dsSimulate(s,'tspan',[0 1000]);

%timer
toc


Produces an error:

Error: Inner matrix dimensions must agree.
in solve_ode_20200402151903_219 (line 59)
in dsSimulate/tryFn (line 1048)
in dsSimulate (line 797)
Error using *
Inner matrix dimensions must agree.

Error in solve_ode_20200402151903_219 (line 59)

I1_E1_iAMPA_IAMPA(1,:)=-I1_E1_iAMPA_s(1,:)*(p.I1_E1_iAMPA_gAMPA*I1_E1_iAMPA_netcon).*(I1_v(1,:)-p.I1_E1_iAMPA_EAMPA);

Error in dsSimulate/tryFn (line 1048)
[outputs{1:length(output_variables)}]=feval(fname);

Error in dsSimulate (line 797)
tryFn(nargout)


Is there a way of setting a matrix/array of values for those connection
parameters just like you can modify the netcon?

[Salva Ardid]

Actually disregard what I said before, I don't think it is correct.

On second thoughts, I think what you need to do is reversing the order, so that the matrix multiplication between s and netcon is done first ( I1_E1_iAMPA_s(1,:)*I1_E1_iAMPA_netcon ), after that the result of this operation should have the same dimensions of that of the conductance and driving force, so just use .* to scale according to conductance:

- p.I1_E1_iAMPA_gAMPA.* ( I1_E1_iAMPA_s(1,:)*I1_E1_iAMPA_netcon ) .* (driving force)

If that is not what you aim, think about inserting the scaling factor in the connectivity netcon

Hope it helps,
Salva

El dijous, 2 d’abril de 2020, a les 15:45:52 EDT, Salva Ardid va escriure: