parseRxnEqu

 parseRxnEqu
   Gets all metabolite names from a cell array of equations

   metabolites=parseRxnEqu(equations)

   equations     A cell array with equation strings

   metabolites   A cell array with the involved metabolites

   The equations should be written like:
   1 A + 3 B (=> or <=>) 5C + 2 D

   If the equation is expressed as for example '... + (n-1) starch' then
   '(n-1) starch' will be interpreted as one metabolite

   Usage: metabolites=parseRxnEqu(equations)

0001 function metabolites=parseRxnEqu(equations)
0002 % parseRxnEqu
0003 %   Gets all metabolite names from a cell array of equations
0004 %
0005 %   metabolites=parseRxnEqu(equations)
0006 %
0007 %   equations     A cell array with equation strings
0008 %
0009 %   metabolites   A cell array with the involved metabolites
0010 %
0011 %   The equations should be written like:
0012 %   1 A + 3 B (=> or <=>) 5C + 2 D
0013 %
0014 %   If the equation is expressed as for example '... + (n-1) starch' then
0015 %   '(n-1) starch' will be interpreted as one metabolite
0016 %
0017 %   Usage: metabolites=parseRxnEqu(equations)
0018 
0019 if ~iscell(equations)
0020     equations={equations};
0021 end
0022 
0023 metabolites={};
0024 
0025 %Replace the the direction arrows and plus signs with a weird character
0026 %that will be used for parsing
0027 equations=strrep(equations,' <=> ', '€');
0028 equations=strrep(equations,' => ', '€');
0029 equations=strrep(equations,' + ', '€');
0030 equations=strtrim(equations);
0031 
0032 for i=1:numel(equations)
0033     %Split each equation in possible metabolites
0034     candidates=regexp(equations{i},'€','split');
0035     
0036     %If the splitting character is at the end (if exchange rxns), then an
0037     %empty string will exist together with the real ones. Remove it
0038     candidates(cellfun(@isempty,candidates))=[];
0039     
0040     %Now remove the potential coefficient before each metabolite
0041     for j=1:numel(candidates)
0042         %If the metabolite has a coefficient it will look as 'number name'
0043         space=strfind(candidates{j},' ');
0044         
0045         if isempty(space)
0046             %Add the metabolite
0047             metabolites=[metabolites;candidates(j)];
0048         else
0049             potNumber=candidates{j}(1:space(1));
0050             %I use str2double here which can't deal with fractions (1/3 glc
0051             %and so on). I do this because I don't want to risk calling
0052             %functions
0053             [~,isNumber]=str2num(potNumber);
0054             
0055             if isNumber==1
0056                 %Remove the coefficient
0057                 metName=candidates{j}(space(1)+1:end);
0058                 metabolites=[metabolites;metName];
0059             else
0060                 %The metabolite name contained spaces
0061                 metabolites=[metabolites;candidates(j)];
0062             end
0063         end
0064     end
0065 end
0066 
0067 metabolites=strtrim(metabolites);
0068 
0069 %Upper/lower case is treated as different names. This should be checked for
0070 %later since it's bad modelling practice
0071 metabolites=unique(metabolites);
0072 
0073 end