Homec4science
Diffusion Research Scripts (Thomas Bolton) (master) Browse

Scripts_Sharing/ISFC/Script_MakeTextFile.m
bdfe7aadccccmaster

Last Change
Raw File

Script_MakeTextFile.m
View Options

%% I want to write a text file with my information
% so here it is !
%% ISFC (no motion regression)
% Indices for each connection
IDX = Summary_final.excursions(:,2);
% Contains the adjacency matrix information for ISFC
A_ISFC = zeros(n_ROI,n_ROI,59);
% Will contain the t and F statistics for all connections
tval_ISFC = nan(n_conn2,1);
fval_ISFC = nan(3,n_conn2);
% Filling the adjacency matrix for all subjects and connections
for j = 1:59
for i = 1:n_conn2
A_ISFC(C1(i),C2(i),j) = n_excursions(i,j);
end
A_ISFC(:,:,j) = A_ISFC(:,:,j) + A_ISFC(:,:,j)';
end
% I will need t values for the plotting (to know what group dominates),
% so I compute them here
for i = 1:n_conn2
[~,~,~,stats] = ttest2(A_ISFC(C1(i),C2(i),Group_Label==1),...
A_ISFC(C1(i),C2(i),Group_Label==2));
% Filling both the matrix and a vector
tval_ISFC(i) = stats.tstat;
clear stats
end
% Now running my ANOVA test...
for i = 1:n_conn2
% Current set of values to consider
tmp_anova = squeeze(A_ISFC(C1(i),C2(i),:));
[~,table] = anovan(tmp_anova,...
{Group_Label' Session_Label'},'model',2,'varnames',...
{'Group_Label','Session_Label'},'display','none');
fval_ISFC(:,i) = [table{2,6};table{3,6};table{4,6}];
clear tmp_anova
clear table
end
% Selects only the TD > ASD connections
f_table1 = fval_ISFC(1,IDX == 1)';
t_table1 = tval_ISFC(IDX==1);
C1_table1 = C1(IDX==1);
C2_table1 = C2(IDX==1);
% Orders
[~,idx_f_table1] = sort(f_table1,'descend');
Table1 = table(C1_table1(idx_f_table1)',C2_table1(idx_f_table1)',round(f_table1(idx_f_table1),2),round(t_table1(idx_f_table1),2));
Table1.Properties.VariableNames = {'Node1','Node2','F','t'};
writetable(Table1,'EXC002_mot_TD_>_ASD.txt','Delimiter','\t');
% Selects only the ASD > TD connections
f_table2 = fval_ISFC(1,IDX == -1)';
t_table2 = tval_ISFC(IDX==-1);
C1_table2 = C1(IDX==-1);
C2_table2 = C2(IDX==-1);
% Orders
[~,idx_f_table2] = sort(f_table2,'descend');
Table2 = table(C1_table2(idx_f_table2)',C2_table2(idx_f_table2)',round(f_table2(idx_f_table2),2),round(t_table2(idx_f_table2),2));
Table2.Properties.VariableNames = {'Node1','Node2','F','t'};
writetable(Table2,'EXC002_mot_ASD_>_TD.txt','Delimiter','\t');
% Now I want a table with only nodes
% Selects only what I need and sets the rest to 0
IDX_TD = IDX;
IDX_TD(isnan(IDX_TD)) = 0;
IDX_TD(IDX_TD ~= 1) = 0;
MAT_TD = jVecToUpperTriMat(IDX_TD,n_ROI)+ jVecToUpperTriMat(IDX_TD,n_ROI)';
DEG_TD = sum(MAT_TD,1);
Nodes = 1:270;
Nodes = Nodes(DEG_TD > 0);
DEG_TD = DEG_TD(DEG_TD > 0);
[~,idx_deg_ISFC_TD] = sort(DEG_TD,'descend');
Table3 = table(Nodes(idx_deg_ISFC_TD)',DEG_TD(idx_deg_ISFC_TD)');
Table3.Properties.VariableNames = {'Node','n_connections'};
writetable(Table3,'EXC002_mot_TD_nodes.txt','Delimiter','\t');
% Now I want a table with only nodes for ASD
% Selects only what I need and sets the rest to 0
IDX_ASD = IDX;
IDX_ASD(isnan(IDX_ASD)) = 0;
IDX_ASD(IDX_ASD ~= -1) = 0;
IDX_ASD = abs(IDX_ASD);
MAT_ASD = jVecToUpperTriMat(IDX_ASD,n_ROI)+ jVecToUpperTriMat(IDX_ASD,n_ROI)';
DEG_ASD = sum(MAT_ASD,1);
Nodes = 1:270;
Nodes = Nodes(DEG_ASD > 0);
DEG_ASD = DEG_ASD(DEG_ASD > 0);
[~,idx_deg_ISFC_ASD] = sort(DEG_ASD,'descend');
Table4 = table(Nodes(idx_deg_ISFC_ASD)',DEG_ASD(idx_deg_ISFC_ASD)');
Table4.Properties.VariableNames = {'Node','n_connections'};
writetable(Table4,'EXC002_mot_ASD_nodes.txt','Delimiter','\t');
CM = zeros(n_conn2,3);
CM2 = zeros(n_conn2,3);
for c = 1:n_conn2
if isnan(IDX(c))
CM(c,:) = [0.8 0.8 0.8];
elseif abs(IDX(c)) ~= 1
CM(c,:) = [0 0 0];
elseif IDX(c) == 1
CM(c,:) = [0.2 0.4 0.8];
else
CM(c,:) = [1 0.2 0.2];
end
end
Size = zeros(n_conn2,1);
Size(isnan(IDX))=0.5;
Size(IDX==1) = 3;
Size(IDX==-1) = 3;
Size(IDX==0) = 1;
Size(abs(IDX)>1) = 1;
figure;
set(gcf,'color','w');
hold on;
grid on
plot([-0.1 0.25],[-0.1 0.25],'color',[0.5 0.5 0.5]);
scatter(mean(ISFC_mot(~isnan(IDX),Group_Label==1),2),mean(ISFC_mot(~isnan(IDX),Group_Label==2),2),Size(~isnan(IDX)),CM(~isnan(IDX),:));
axis([-0.1 0.25 -0.1 0.25]);
DELTA_ISFC = mean(ISFC_mot(:,Group_Label==1),2) - mean(ISFC_mot(:,Group_Label==2),2);
figure;
set(gcf,'color','w');
hold on;
grid on
plot(DELTA_ISFC(abs(IDX)==1),mean(ISFC_mot(abs(IDX)==1,Group_Label==1),2),'b.');
plot(DELTA_ISFC(abs(IDX)==1),mean(ISFC_mot(abs(IDX)==1,Group_Label==2),2),'r.');
corr(mean(ISFC_mot(:,Group_Label==1),2),mean(ISFC_mot(:,Group_Label==2),2))
for i = 1:10000
randidx = randperm(59);
TD = mean(ISFC_mot(:,randidx(1:30)),2);
ASD = mean(ISFC_mot(:,randidx(31:end)),2);
R_null(i) = corr(TD,ASD);
end
c4science · Help