-
Notifications
You must be signed in to change notification settings - Fork 1
/
CalculatingGoodnessOfFitDeactivation.m
237 lines (158 loc) · 5.39 KB
/
CalculatingGoodnessOfFitDeactivation.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
function CalculatingGoodnessOfFitDeactivation(exp_ref)
% Compares predictions of current response to deactivation protocol for experiment exp_ref
% for simulated cell-specific and average models and leading literature models.
% Average model and literature models have their conductance scaled so that they minimise the
% square difference between the simulated model trace and the current in response to the action
% potential waveform for exp_ref.
% Import literature model parameters
cd ../ParameterSets
W=importdata('WangModelSimulatedParameters.mat');
DV=importdata('DiVeroliRTModelSimulatedParameters.mat');
M=importdata('MazhariModelSimulatedParameters.mat');
TT=importdata('TenTusscherModelSimulatedParameters.mat');
Z=importdata('ZengModelSimulatedParameters.mat');
Z_tol=importdata('ZengTolModelSimulatedParameters.mat');
cd ..
% Identify temperature for experiment
if strcmp(exp_ref,'16708016')==1
temperature = 21.8;
end
if strcmp(exp_ref,'16708060')==1
temperature = 21.7;
end
if strcmp(exp_ref,'16704047')==1
temperature = 21.6;
end
if strcmp(exp_ref,'16704007')==1
temperature = 21.2;
end
if strcmp(exp_ref,'16713003')==1
temperature = 21.3;
end
if strcmp(exp_ref,'16713110')==1||strcmp(exp_ref,'16715049')==1
temperature = 21.4;
end
if strcmp(exp_ref,'16707014')==1
temperature = 21.4;
end
if strcmp(exp_ref,'16708118')==1
temperature = 21.7;
end
% Imports deactivation, sine wave and action potential protocols
cd Protocols
V=importdata('deactivation_protocol.mat');
V_AP=importdata('ap_protocol.mat');
V_sine = importdata(['sine_wave_protocol.mat']);
cd ..
cd ExperimentalData
cd(exp_ref)
D=importdata(['deactivation_',exp_ref,'_dofetilide_subtracted_leak_subtracted.mat']);
D_AP=importdata(['ap_',exp_ref,'_dofetilide_subtracted_leak_subtracted.mat']);
cd ..
cd ..
cd Code
% Identifies maximum likelihood parameters for cell-specific and average model for fit to sine wave
[chain,likelihood] = FindingBestFitsAfterMCMC('hh',{'sine_wave'},exp_ref);
[i,v]= max(likelihood);
k= chain(v,:);
[chain_av,likelihood_av] = FindingBestFitsAfterMCMC('hh',{'sine_wave'},'average');
[i,v]= max(likelihood_av);
k_av= chain_av(v,:);
% Simulate sine wave protocol for average model and scale averae model conductance to minimise square difference
% between average model and cell-specific experimental recording of sine wave.
I=SimulatingData(35,{'deactivation'},k,V,temperature);
I_av=SimulatingData(35,{'deactivation'},k_av,V,temperature);
I_av_sine = SimulatingData(35,{'sine_wave'},k_av,V_sine,temperature);
J=I;
F=D;
cd ../ExperimentalData
cd(exp_ref)
cell_specific_sine_wave=importdata(['sine_wave_',exp_ref,'_dofetilide_subtracted_leak_subtracted.mat']);
cd ..
cd ..
c=0;
for a = 0.000001:0.01:5
c=c+1;
Diff(1,c) = sum((cell_specific_sine_wave-a.*I_av_sine).^2)./length(I_av_sine);
end
[i,v]=min((Diff(1,:)));
s_av = v*0.01+0.000001;
J_av = s_av.*I_av;
I_av=J_av;
K_av = I_av;
cd Code
% Scale literature models conductance to minimise square difference between simulation of action potential
% and experimental recording for exp_ref of action potential.
% Wang model
IW = SimulatingData(27,{'deactivation'},W,V,temperature);
IW_AP = SimulatingData(27,{'ap'},W,V_AP,temperature);
c=0;
for a = 0.000001:0.01:5
c=c+1;
Diff(1,c) = sum((D_AP-a.*IW_AP).^2)./length(IW_AP);
end
[i,v]=min((Diff(1,:)));
sw = v*0.01+0.000001;
JW = sw.*IW;
IW=JW;
%DiVeroli Model
ID = SimulatingData(5,{'deactivation'},DV,V,temperature);
ID_AP = SimulatingData(5,{'ap'},DV,V_AP,temperature);
c=0;
for a = 0.000001:0.01:5
c=c+1;
Diff(1,c) = sum((D_AP-a.*ID_AP).^2)./length(ID_AP);
end
[i,v]=min((Diff(1,:)));
sd = v*0.01+0.000001;
JD = sd.*ID;
ID=JD;
%Mazhari model
IM = SimulatingData(16,{'deactivation'},M,V,temperature);
IM_AP = SimulatingData(16,{'ap'},M,V_AP,temperature);
c=0;
for a = 0.000001:0.01:5
c=c+1;
Diff(1,c) = sum((D_AP-a.*IM_AP).^2)./length(IM_AP);
end
[i,v]=min((Diff(1,:)));
sm = v*0.01+0.000001;
JM = sm.*IM;
IM=JM;
% TenTusscher model
IT = SimulatingData(26,{'deactivation'},TT,V,temperature);
IT_AP = SimulatingData(26,{'ap'},TT,V_AP,temperature);
c=0;
for a = 0.000001:0.01:5
c=c+1;
Diff(1,c) = sum((D_AP-a.*IT_AP).^2)./length(IT_AP);
end
[i,v]=min((Diff(1,:)));
st = v*0.01+0.000001;
JT = st.*IT;
IT=JT;
%Zeng model
IZ = SimulatingData(29,{'deactivation'},Z,V,temperature);
IZ_AP = SimulatingData(2999,{'ap'},Z_tol,V_AP,temperature);
c=0;
for a = 0.000001:0.01:5
c=c+1;
Diff(1,c) = sum((D_AP-a.*IZ_AP).^2)./length(IZ_AP);
end
[i,v]=min((Diff(1,:)));
sz = v*0.01+0.000001;
JZ = sz.*IZ;
IZ=JZ;
JJZ =IZ;
JJT = IT;
JJW = IW;
JJM = IM;
JJD = ID;
% Calculate differences between literature, cell specific and average models and deactivation experiment
cell_specific_diff = sum((sqrt((F-J).^2)))/length(J)
average_diff = sum((sqrt((F-K_av).^2)))/length(K_av)
wang_diff = sum((sqrt((F-JJW).^2)))/length(JJW)
diveroli_diff = sum((sqrt((F-JJD).^2)))/length(JJD)
mazhari_diff = sum((sqrt((F-JJM).^2)))/length(JJM)
tentusscher_diff = sum((sqrt((F-JJT).^2)))/length(JJT)
zeng_diff = sum((sqrt((F-JJZ).^2)))/length(JJZ)