tic
N=21000; %ms
trials=2;
dt=0.05; %ms
n_spike=0; % current number of spikes
sum=0;
cormax=2; % correlation time
% cell 1
c=268;
gl=8.47;
el=-51.31;
vt=-52.23;
delta=0.84;
vreset=-68;
a=37.79; tauw=20.76; b=441;
% dendritic filtering
tauc=0.317; % ms
taus=30.91; % ms2
c1=66.97; % microF/cm^2
cm=1e6; % microF
S1=c1/cm; % S_dend, calculated acccording "pulse" fit
p=0.065; % S_soma / (S_soma + S_dend)
gc=cm*p*(1-p)*(1/tauc-1/taus); % coupling conductance
% gc=0;
G=gl +gc/p/(1-p); % total conductance: leak conductance + coupling conductance;
for j=1:1:trials % trials
Ihold=-150;
sigma=0;
corr=cormax;
n_spike=0;
vspike=0;
temp=0;
% initial conditions
v(1)=-60;
w(1)=-300;
input(1)=Ihold;
V(1)=0; % V=vs-vd; vs=vd at t=0
t(1)=0;
m(1)=0;
in(1)=0;
% parameters for external biexponential intput
Am=0; % pA, optimal for inhibition
taus1=1.5; % ms, rise constant
taus2=10; % ms, decay constant
ts=500; % STIMULUS TIME!
% bassin of attraction
vb=importdata('vb-150.mat');
wb=importdata('wb-150.mat');
%down=0; % auxilary variable
tt=0;
td=1000; % time of one noise/silent period
n=0;
q=0;
for i=2:1:round(N/dt)
t(i)=(i-1)*dt;
% increasing noise steps
if t(i)>=td
n=n+1;
sigma=n*5; % noise amplitude on the each step
% FR(n)=n_spike/1; % record the rate during the noise stimulation, Hz
% SG(n)=sigma;
if mod(td/1000,2)==0 % zero variance, "pauses"
q=q+1; % counter for increasing noises
FR(j,q)=n_spike/1; % record the rate during the noise stimulation, Hz
SG(q)=sigma;
sigma=0;
end
td=td+1000; % next period for noise
n_spike=0; % reset the spike number after each sigma change
end
% Sarah stimulus
%{
%if t(i)>600
% Ihold=-1400;
%end
%if t(i)>601
% Ihold=-400;
%end
%if t(i)>800
% Ihold=-700;
%end
%if t(i)>1300
% Ihold=-400;
%end
%}
% GENERATE EXTERNAL STIMULI
% delta function approximation
%{
if t(i)==ts
stim=1/dt;
% new stimulus time for randomness
ts=ts + ts;
else
stim=0;
end;
m(i)=dt/taus1/taus2*(Am*(1-in(i-1))*stim/K(1/taus1,1/taus2)-in(i-1)-(taus1+taus2)*m(i-1)) + m(i-1);
in(i)=m(i)*dt + in(i-1);
%}
temp=temp-dt/corr*temp + sqrt(2*dt/corr)*randn(1,1);
% no compensation
input(i)=Ihold + temp*sigma;
%+ in(i);
%{
% compensation
% input(i)=1/(1-gc/p/G)*(Ihold +temp*sigma);
% dendrite aproximation
% V=v-vd, applying the subthreshould approximation + mean(gc/p*V)
% v(i)=dt/c*( -gl*(v(i-1)-el) +gl*delta*exp((v(i-1)-vt)/delta) -w(i-1) + input(i) -gc/p*V(i-1) ) + v(i-1);
% V(i)=dt/c*( -G*V(i-1) +input(i) ) + V(i-1);
% w(i)=dt/tauw*(a*(v(i-1)-el)-w(i-1)) + w(i-1);
%}
% no dendrite
v(i)=dt/c*(-gl*(v(i-1)-el)+gl*delta*exp((v(i-1)-vt)/delta)-w(i-1)+input(i) ) + v(i-1);
w(i)=dt/tauw*(a*(v(i-1)-el)-w(i-1)) + w(i-1);
if v(i)>=vspike
n_spike=n_spike+1;
v(i-1)=0; % add sticks to the previous step
v(i)=vreset;
w(i)=w(i) + b;
end
% if inpolygon(v(i),w(i),vb,wb) == 1 % time inside of the attraction bassin
% tt=tt+1;
% time=tt*dt;
% end
end
end % measuring of the averaged response
%% Plot
subplot(3,1,3)
plot(t,v);
xlabel('time, ms');
ylabel('V, mV');
title('Increasing noise variances');
subplot(3,1,2)
plot(t,input);
xlabel('time, ms');
ylabel('In, pA');
[SG_sorted, SG_ind] = sort([0,SG]);
FR_all=[22,mean(FR)];
FR_std=[0,std(FR)];
FR_sorted = FR_all(SG_ind);
FR_std_sorted=FR_std(SG_ind);
% +one more 22Hz point without noise
subplot(3,1,1)
%errorbar([0,SG],[22,mean(FR)],[0,std(FR)],'.');
errorbar(SG_sorted,FR_sorted,FR_std_sorted,'.');
xlabel('Sigma, pA');
ylabel('Rate, Hz');
%%
toc