extensions of v2

Dependencies/FID-A-MASTER/io_loadlcmdetail.m

@@ -0,0 +1,140 @@
+%io_loadlcmdetail.m
+%Jamie Near, McGill University 2014.
+%
+% USAGE:
+% [metabs,corrMatrix]=io_loadlcmdetail(filename);
+% 
+% DESCRIPTION:
+% This function loads in the "detailed output" of LCModel and returns the
+% matrix of metabolite correlation coefficients.
+% 
+% INPUTS:
+% filename   = Filename of the lcmodel detailed output file.  
+
+function [metabs,corrMatrix]=io_loadlcmdetail(filename);
+
+%First open the detailed output file and search for the beginning of the
+%Correlation Coefficients table
+fid=fopen(filename);
+line=fgets(fid);
+search_index=findstr(line,'Correlation coefficients');
+
+while isempty(search_index)
+    line=fgets(fid);
+    search_index=findstr(line,'Correlation coefficients');
+end
+
+%Now that we've found the top of the correlation coefficients table, we
+%need to record the names of all of the metabolites that are listed in the
+%table.  Start by going down two lines to where the metabolite names are
+%listed:
+line=fgets(fid);
+line=fgets(fid);
+
+%Read the first line of metabolite names and then go down one line.  The 
+%table of correlation coefficients can only fit 17 metabolites per row.
+%Therefore if there are more than 17 metabolites to read, then we will have
+%to go down another row and read again.  A simple trick to know whether to
+%go down another row is to check and see if the first token in the next row
+%is equal to the second token from the first row.  If so, then we're done
+%reading metabolite names.  If not, then we need to keep reading.  
+metabstr=line;
+[temp,remain]=strtok(line);
+secondMetabName=strtok(remain);
+line=fgets(fid);
+
+while ~strcmp(strtok(line),secondMetabName)
+    metabstr=[metabstr line];
+    line=fgets(fid);
+end
+
+%We are now finished reading all of the metabolite names.  Now make a cell 
+%array containing the names of all of the meatbolites.  Note that at this
+%point in the code, the final element in metabs will be a blank space,
+%since the last metabolite name is missing from the top row of the correlation
+%coefficient table in the LCModel detailed output.
+remain=metabstr;
+n=1;
+while ~isempty(remain)
+    [metabs{n},remain]=strtok(remain);
+    n=n+1;
+end
+
+%Now calculate the total number of metabolites, and initialize the 
+%correlation coefficient matrix:
+numMetabs=numel(metabs);
+corrMatrix=zeros(numMetabs);
+fillLine=1;
+
+%Now start populating the correlation coefficient matrix.  The first token
+%of each line is a metabolite name, so we have to remove it.  The rest of
+%the line is a row in the correlation matrix.  The first 17 lines are
+%simple to read becuase they all have 17 elements or less, so they don't
+%wrap around the page (remember there are only 17 columns in the page).  NOTE:
+%The correlation matrix is symmetric about the diagonal, but we start by only
+%constructing the lower half of the diagonal (The upper half will be all 
+%zeros).  The diagonal elements themselves will be set to 0.5.  Then later, 
+%the full matrix, will be completed by adding the matrix to it's transpose,
+%thus completing the symmetry, and making all of the diagonals ones.  We
+%will also rename "metabs" as we go, becuase the full names of each
+%metabolite are given with greater precision on the left column of the
+%table in the lcmodel detailed output (compared to the top row of the table).
+corrMatrix(1,1)=0.5;
+if numMetabs>17
+    upto=18;
+else
+    upto=numMetabs;
+end
+while fillLine<upto
+    [name,corrs]=strtok(line);
+    metabs{fillLine+1}=name;
+    corrs=[corrs(1:end-1) ' 0.5'];
+    corrMatrix(fillLine+1,1:length(str2num(corrs)))=str2num(corrs);
+    line=fgets(fid);
+    fillLine=fillLine+1;
+end
+
+
+
+line=fgets(fid);
+
+%For the 18th to 34th lines, there is one line of 'wrap around'.
+if numMetabs>34
+    upto=35;
+else
+    upto=numMetabs;
+end
+while fillLine<upto
+    [name,corrs]=strtok(line);
+    metabs{fillLine+1}=name;
+    line=fgets(fid);
+    corrs=[corrs(1:end-1) ' ' line(1:end-1) ' 0.5'];
+    corrMatrix(fillLine+1,1:length(str2num(corrs)))=str2num(corrs);
+    line=fgets(fid);
+    fillLine=fillLine+1;
+end
+
+
+%For the 35th to 52nd lines, there are two lines of 'wrap around'.  Assume
+%that there will be no more than 52 elemens in the correlation matrix.
+while fillLine<numMetabs
+    [name,corrs]=strtok(line);
+    metabs{fillLine+1}=name;
+    line1=fgets(fid);
+    line2=fgets(fid);
+    corrs=[corrs(1:end-1) ' ' line1(1:end-1) ' ' line2(1:end-1) ' 0.5'];
+    size(str2num(corrs));
+    corrMatrix(fillLine+1,1:length(str2num(corrs)))=str2num(corrs);
+    line=fgets(fid);
+    fillLine=fillLine+1;
+end
+metabs{end}=name;
+
+%Now add the correlation matrix (the lower have of which is currently only
+%populated) to its transpose to complete the symmetry.
+corrMatrix=corrMatrix+corrMatrix';
+
+fclose(fid);
+
+% imagesc(corrMatrix);
+% impixelinfo;

Dependencies/FID-A-MASTER/io_readlcmcoord.m

@@ -0,0 +1,74 @@
+%io_readlcmcoord.m
+%Jamie Near, McGill University 2014.
+%
+% USAGE:
+% out = io_readlcmcoord(filename,metab) 
+% 
+% DESCRIPTION:
+% Reads a LCModel .coord file and extracts the desired part.
+% 
+% INPUTS:
+% filename   = filename of the LCModel .coord file.
+% part       = Which metabolite fit to extract from the .coord file - 
+%              The abbreviated metabolite name should be given (ie.
+%              'Cr','PCr','Glu','GABA',etc.)
+% 
+%Esin Ozturk Isik, Bogazici University, 01/06/2017.  Corrected the handling
+%of non-existing metabolites.
+
+function [out]=io_readlcmcoord(filename,metab)
+
+fid=fopen(filename);
+linenum=1;
+line=fgets(fid);
+
+ppmstr='points on ppm-axis =';
+ppm_index=findstr(line,ppmstr);
+while isempty(ppm_index)
+    line=fgets(fid);
+    ppm_index=findstr(line,ppmstr);
+end
+
+Npts=str2num(line(2:5));
+
+line=fgets(fid);
+while linenum<=(Npts/10)
+    [A,count, errmsg, nextindex] = sscanf(line, '%f', inf);
+    A;
+    ppm((linenum-1)*10+1:linenum*10,1)=A;
+    linenum=linenum+1;
+    line=fgets(fid);
+end
+
+
+metab_str='';
+metab_str(2:1+length(metab))=metab;
+metab_str(11:17)='Conc. =';
+metab_str;
+metab_index=findstr(line,metab);
+while isempty(metab_index) & ~feof(fid)
+    line=fgets(fid);
+
+    metab_index=findstr(line,metab);
+end
+linenum=1;
+line=fgets(fid);
+
+if ~feof(fid)
+while linenum<=(Npts/10) 
+    [A,count, errmsg, nextindex] = sscanf(line, '%f', inf);
+    A;
+    specs((linenum-1)*10+1:linenum*10,1)=A;
+    linenum=linenum+1;
+    line=fgets(fid);
+end
+
+out.ppm=ppm;
+out.specs=specs;
+else
+out.ppm=[];
+out.specs=[];
+fclose('all');
+end
+
+

Dependencies/FID-A-MASTER/io_readlcmcoord_getBackground.m

@@ -0,0 +1,74 @@
+%io_readlcmcoord_getBackground.m
+%Jamie Near, McGill University 2014.
+%
+% USAGE:
+% out = io_readlcmcoord_getBackground(filename,part) 
+% 
+% DESCRIPTION:
+% Reads a LCModel .coord file and extracts the desired part.
+% 
+% INPUTS:
+% filename   = filename of the LCModel .coord file.
+% part       = Which part of the .coord file to extract - 'bg' extracts the
+%             LCModel baseline signal, 'sp' extracts the spectrum, and 
+%             'fit' extracts the fit.
+
+function [out]=io_readlcmcoord_getBackground(filename,part)
+
+fid=fopen(filename);
+linenum=1;
+line=fgets(fid)
+
+
+%READ PPM AXIS
+ppmstr='points on ppm-axis ='
+ppm_index=findstr(line,ppmstr);
+while isempty(ppm_index)
+    line=fgets(fid);
+    ppm_index=findstr(line,ppmstr);
+end
+
+Npts=str2num(line(2:5))
+
+line=fgets(fid);
+while linenum<=(Npts/10)
+    [A,count, errmsg, nextindex] = sscanf(line, '%f', inf);
+    A;
+    ppm((linenum-1)*10+1:linenum*10,1)=A;
+    linenum=linenum+1;
+    line=fgets(fid);
+end
+
+
+
+%GET BACKGROUND
+
+linenum=1;
+switch part
+    case 'bg'
+        bgstr='NY background values follow'
+    case 'sp'
+        bgstr='NY phased data points follow'
+    case 'fit'
+        bgstr='NY points of the fit to the data follow'
+    otherwise
+        error('ERROR:  part not found');
+end
+bg_index=findstr(line,bgstr);
+while isempty(bg_index)
+    line=fgets(fid);
+    bg_index=findstr(line,bgstr);
+end
+
+line=fgets(fid);
+while linenum<=(Npts/10)
+    [A,count, errmsg, nextindex] = sscanf(line, '%f', inf);
+    A;
+    bg((linenum-1)*10+1:linenum*10,1)=A;
+    linenum=linenum+1;
+    line=fgets(fid);
+end
+
+
+out.ppm=ppm;
+out.specs=bg;