Merge pull request #29 from johan-gson/master

Adapted the HPA code to be able to import the tsv files published for…

Author: Simonas Marcišauskas

hpa/parseHPA.m

@@ -1,13 +1,18 @@
-function hpaData=parseHPA(fileName)
+function hpaData=parseHPA(fileName, version)
 % parseHPA
 %   Parses a database dump of the Human Protein Atlas (HPA)
 %
-%   fileName            comma-separated database dump of HPA. For details
+%   fileName            comma- or tab-separated database dump of HPA. For details
 %                       regarding the format, see
 %                       http://www.proteinatlas.org/about/download.
+%   version             version of HPA [optional, default=17]
 %
 %   hpaData
-%       genes               cell array with the unique gene names
+%       genes               cell array with the unique gene names. In
+%                           version 17 this is the ensamble name, see
+%                           geneNames below for the names in ver 17
+%       geneNames           cell array with the gene names, indexed the
+%                           same way as genes.
 %       tissues             cell array with the tissue names. The list may not be
 %                           unique, as there can be multiple cell types per tissue
 %       celltypes           cell array with the cell type names for each tissue
@@ -20,7 +25,8 @@
 %                           hpaData.levels of gene i in cell type j
 %       gene2Type           gene-to-evidence type mapping in sparse matrix form.
 %                           The value for element i,j is the index in
-%                           hpaData.types of gene i in cell type j
+%                           hpaData.types of gene i in cell type j. Doesn't
+%                           exist in version 17.
 %       gene2Reliability    gene-to-reliability level mapping in sparse matrix form.
 %                           The value for element i,j is the index in
 %                           hpaData.reliabilities of gene i in cell type j
@@ -29,33 +35,67 @@
 %   Usage: hpaData=parseHPA(fileName)
 %
 %   Rasmus Agren, 2014-01-08
-%
+%   Johan Gustafsson, 2017-10-10
+
+if nargin<2
+    version=17; %Change this and add code for more versions when the current HPA version is increased and the format is changed
+end;
+
+if (version == 17)
+    fid=fopen(fileName,'r');
+    hpa=textscan(fid,'%q %q %q %q %q %q','Delimiter','\t');
+    fclose(fid);
+
+    %Go through and see if the headers match what was expected
+    headers={'Gene' 'Gene name' 'Tissue' 'Cell type' 'Level' 'Reliability'};
+    for i=1:numel(headers)
+        if ~strcmpi(headers(i),hpa{i}(1))
+            EM=['Could not find the header "' headers{i} '". Make sure that the input file matches the format specified at http://www.proteinatlas.org/about/download'];
+          dispEM(EM);
+        end
+        %Remove the header line here
+        hpa{i}(1)=[];
+    end
+
+    %Get the unique values of each data type
+    [hpaData.genes, P, I]=unique(hpa{1});
+    hpaData.geneNames=hpa{2}(P); %make this vector use the index as genes
+    [~, J, K]=unique(strcat(hpa{3},'€',hpa{4}));
+    hpaData.tissues=hpa{3}(J);
+    hpaData.celltypes=hpa{4}(J);
+    [hpaData.levels, ~, L]=unique(hpa{5});
+    [hpaData.reliabilities, ~, N]=unique(hpa{6});
 
-fid=fopen(fileName,'r');
-hpa=textscan(fid,'%q %q %q %q %q %q','Delimiter',',');
-fclose(fid);
+    %Map the data to be sparse matrises instead
+    hpaData.gene2Level=sparse(I,K,L,numel(hpaData.genes),numel(hpaData.tissues));
+    hpaData.gene2Reliability=sparse(I,K,N,numel(hpaData.genes),numel(hpaData.tissues));
+else
+    fid=fopen(fileName,'r');
+    hpa=textscan(fid,'%q %q %q %q %q %q','Delimiter',',');
+    fclose(fid);
 
-%Go through and see if the headers match what was expected
-headers={'Gene' 'Tissue' 'Cell type' 'Level' 'Expression type' 'Reliability'};
-for i=1:numel(headers)
-    if ~strcmpi(headers(i),hpa{i}(1))
-    	EM=['Could not find the header "' headers{i} '". Make sure that the input file matches the format specified at http://www.proteinatlas.org/about/download'];
-      dispEM(EM);
+    %Go through and see if the headers match what was expected
+    headers={'Gene' 'Tissue' 'Cell type' 'Level' 'Expression type' 'Reliability'};
+    for i=1:numel(headers)
+        if ~strcmpi(headers(i),hpa{i}(1))
+            EM=['Could not find the header "' headers{i} '". Make sure that the input file matches the format specified at http://www.proteinatlas.org/about/download'];
+          dispEM(EM);
+        end
+        %Remove the header line here
+        hpa{i}(1)=[];
     end
-    %Remove the header line here
-    hpa{i}(1)=[];
-end
-
-%Get the unique values of each data type
-[hpaData.genes, ~, I]=unique(hpa{1});
-[~, J, K]=unique(strcat(hpa{2},'€',hpa{3}));
-hpaData.tissues=hpa{2}(J);
-hpaData.celltypes=hpa{3}(J);
-[hpaData.levels, ~, L]=unique(hpa{4});
-[hpaData.types, ~, M]=unique(hpa{5});
-[hpaData.reliabilities, ~, N]=unique(hpa{6});
-
-%Map the data to be sparse matrises instead
-hpaData.gene2Level=sparse(I,K,L,numel(hpaData.genes),numel(hpaData.tissues));
-hpaData.gene2Type=sparse(I,K,M,numel(hpaData.genes),numel(hpaData.tissues));
-hpaData.gene2Reliability=sparse(I,K,N,numel(hpaData.genes),numel(hpaData.tissues));
+
+    %Get the unique values of each data type
+    [hpaData.genes, ~, I]=unique(hpa{1});
+    [~, J, K]=unique(strcat(hpa{2},'€',hpa{3}));
+    hpaData.tissues=hpa{2}(J);
+    hpaData.celltypes=hpa{3}(J);
+    [hpaData.levels, ~, L]=unique(hpa{4});
+    [hpaData.types, ~, M]=unique(hpa{5});
+    [hpaData.reliabilities, ~, N]=unique(hpa{6});
+
+    %Map the data to be sparse matrises instead
+    hpaData.gene2Level=sparse(I,K,L,numel(hpaData.genes),numel(hpaData.tissues));
+    hpaData.gene2Type=sparse(I,K,M,numel(hpaData.genes),numel(hpaData.tissues));
+    hpaData.gene2Reliability=sparse(I,K,N,numel(hpaData.genes),numel(hpaData.tissues));    
+end