Mercurial > repos > insilico-bob > mda_heatmap_gen

changeset 0:781d999e67b5 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded uses local dir to access sh and java file
author insilico-bob
date Thu, 07 Apr 2016 11:40:48 -0400
parents
children ec2453936196
files ._mda_heatmap_gen mda_heatmap_gen/._.DS_Store mda_heatmap_gen/._mda_heatmap_gen.xml mda_heatmap_gen/CHM.R mda_heatmap_gen/GalaxyMapGen.jar mda_heatmap_gen/heatmap.sh mda_heatmap_gen/mda_heatmap_gen.xml
diffstat 7 files changed, 240 insertions(+), 0 deletions(-) [+]
line wrap: on
line diff
Binary file ._mda_heatmap_gen has changed
Binary file mda_heatmap_gen/._.DS_Store has changed
Binary file mda_heatmap_gen/._mda_heatmap_gen.xml has changed
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/mda_heatmap_gen/CHM.R	Thu Apr 07 11:40:48 2016 -0400
@@ -0,0 +1,125 @@
+### This method generates a row and column ordering given an input matrix and ordering methods.
+###
+### matrixData - numeric matrix 
+### rowOrderMethod - Hierarchical, Original, Random
+### rowDistanceMeasure - For clustering, distance measure. May be: euclidean, binary, manhattan, maximum, canberra, minkowski, or correlation.
+### rowAgglomerationMethod - For clustering, agglomeration method.  May be:  'average' for Average Linkage, 'complete' for Complete Linkage,
+###                                                                          'single' for Single Linkage, 'ward', 'mcquitty', 'median', or 'centroid'.
+### colOrderMethod 
+### colDistanceMeasure
+### colAgglomerationMethod
+### rowOrderFile - output file of order of rows 
+### rowDendroFile - output file of row dendrogram  
+### colOrderFile - output file of order of cols
+### colDendroFile - output file of col dendrogram
+
+performDataOrdering<-function(dataFile, rowOrderMethod, rowDistanceMeasure, rowAgglomerationMethod, colOrderMethod, colDistanceMeasure, colAgglomerationMethod,rowOrderFile, colOrderFile, rowDendroFile, colDendroFile)
+{ 
+   dataMatrix = read.table(dataFile, header=TRUE, sep = "\t", row.names = 1, as.is=TRUE)
+   rowOrder <-  createOrdering(dataMatrix, rowOrderMethod, "row", rowDistanceMeasure, rowAgglomerationMethod)  
+   if (rowOrderMethod == "Hierarchical") {
+      writeHCDataTSVs(rowOrder, rowDendroFile, rowOrderFile)
+   } else {
+      writeOrderTSV(rowOrder, rownames(dataMatrix), rowOrderFile)
+   }
+
+   colOrder <-  createOrdering(dataMatrix, colOrderMethod, "col", colDistanceMeasure, colAgglomerationMethod)  
+   if (colOrderMethod == "Hierarchical") {
+      writeHCDataTSVs(colOrder, colDendroFile, colOrderFile)
+   } else {
+      writeOrderTSV(colOrder, colnames(dataMatrix), colOrderFile)
+   }
+}
+
+#creates output files for hclust ordering
+writeHCDataTSVs<-function(uDend, outputHCDataFileName, outputHCOrderFileName)
+{
+   data<-cbind(uDend$merge, uDend$height, deparse.level=0)
+   colnames(data)<-c("A", "B", "Height")
+   write.table(data, file = outputHCDataFileName, append = FALSE, quote = FALSE, sep = "\t", row.names=FALSE)
+ 
+   data=matrix(,length(uDend$labels),2);
+   for (i in 1:length(uDend$labels)) {
+      data[i,1] = uDend$labels[i];
+      data[i,2] = which(uDend$order==i);
+   }
+   colnames(data)<-c("Id", "Order")
+   write.table(data, file = outputHCOrderFileName, append = FALSE, quote = FALSE, sep = "\t", row.names=FALSE)
+}
+
+#creates order file for non-clustering methods
+writeOrderTSV<-function(newOrder, originalOrder, outputHCOrderFileName)
+{
+   data=matrix(,length(originalOrder),2);
+   for (i in 1:length(originalOrder)) {
+      data[i,1] = originalOrder[i];
+      data[i,2] = which(newOrder==originalOrder[i]);
+   }
+   colnames(data)<-c("Id", "Order")
+   write.table(data, file = outputHCOrderFileName, append = FALSE, quote = FALSE, sep = "\t", row.names=FALSE)
+}
+
+
+
+createOrdering<-function(matrixData, orderMethod, direction, distanceMeasure, agglomerationMethod)
+{
+  ordering <- NULL
+
+  if (orderMethod == "Hierarchical")
+  {
+
+    # Compute dendrogram for "Distance Metric"
+    distVals <- NULL
+    if(direction=="row") {
+      if (distanceMeasure == "correlation") {
+        geneGeneCor <- cor(t(matrixData), use="pairwise")
+        distVals <- as.dist((1-geneGeneCor)/2)
+      } else {
+        distVals <- dist(matrixData, method=distanceMeasure)
+      }
+    } else { #column
+      if (distanceMeasure == "correlation") {
+        geneGeneCor <- cor(matrixData, use="pairwise")
+        distVals <- as.dist((1-geneGeneCor)/2)
+      } else {
+        distVals <- dist(t(matrixData), method=distanceMeasure)
+      }
+    }
+
+    if (agglomerationMethod == "ward") {
+      ordering <- hclust(distVals * distVals, method="ward.D2")
+    } else {
+      ordering <- hclust(distVals, method=agglomerationMethod)
+    }
+  }
+  else if (orderMethod == "Random")
+  {
+    if(direction=="row") {
+       headerList <- rownames(matrixData)
+       ordering <- sample(headerList, length(headerList)) 
+    } else {
+       headerList <- colnames(matrixData)
+       ordering <- sample(headerList, length(headerList)) 
+    }
+  }
+  else if (orderMethod == "Original")
+  {
+    if(direction=="row") {
+       ordering <- rownames(matrixData) 
+    } else {
+       ordering <- colnames(matrixData) 
+    }
+  } else {
+    stop("createOrdering -- failed to find ordering method")
+  }
+  return(ordering)
+}
+### Initialize command line arguments and call performDataOrdering
+
+options(warn=-1)
+
+args = commandArgs(TRUE)
+
+performDataOrdering(dataFile=args[1], rowOrderMethod=args[2], rowDistanceMeasure=args[3], rowAgglomerationMethod=args[4], colOrderMethod=args[5], colDistanceMeasure=args[6], colAgglomerationMethod=args[7],rowOrderFile=args[8], colOrderFile=args[9], rowDendroFile=args[10], colDendroFile=args[11])
+
+#suppressWarnings(performDataOrdering(dataFile=args[1], rowOrderMethod=args[2], rowDistanceMeasure=args[3], rowAgglomerationMethod=args[4], colOrderMethod=args[5], colDistanceMeasure=args[6], colAgglomerationMethod=args[7],rowOrderFile=args[8], colOrderFile=args[9], rowDendroFile=args[10], colDendroFile=args[11]))
Binary file mda_heatmap_gen/GalaxyMapGen.jar has changed
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/mda_heatmap_gen/heatmap.sh	Thu Apr 07 11:40:48 2016 -0400
@@ -0,0 +1,30 @@
+echo $1 $2 $3 $4 $5 $6 $7 $8 $9 ${10} ${11} ${12} ${13} ${14} ${15}
+
+#create temp directory for row and col order and dendro files.
+tdir=${11}/$(date +%s)
+echo $tdir
+mkdir $tdir
+
+#run R to cluster matrix
+output="$(R --slave --vanilla --file=${11}/CHM.R --args $3 $4 $5 $6 $7 $8 $9 $tdir/ROfile.txt $tdir/COfile.txt $tdir/RDfile.txt $tdir/CDfile.txt)"
+rc=$?; if [[ $rc != 0 ]]; then echo $output; exit $rc; fi
+
+
+#there are a variable number of triplicate parameters for classification bars
+count=0
+classifcations=''
+
+for i in "$@"; do
+  if [ $count -gt 10 ]
+  then
+    classifications=$classifications' '$i
+  fi
+
+  count=$((count+1))
+done
+
+#call java program to generate NGCHM viewer files.
+java -jar ${11}/GalaxyMapGen.jar $1 $2 DataLayer1 $3 linear Row Column $4 $5 $6 $tdir/ROfile.txt $tdir/RDfile.txt $7 $8 $9 $tdir/COfile.txt $tdir/CDfile.txt ${10} $classifications
+
+#clean up tempdir
+rm -rf $tdir
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/mda_heatmap_gen/mda_heatmap_gen.xml	Thu Apr 07 11:40:48 2016 -0400
@@ -0,0 +1,85 @@
+<?xml version="1.0" encoding="UTF-8" ?>
+<tool id="mda_heatmap_gen" name="MDA Heatmap Generation" version="5.2.1">
+  <description>Create HeatMap from Dendogram and Matrix</description>
+<!-- <command>$GALAXY_ROOT_DIR/tools/MDA_Heatmaps/heatmap.sh  '$hmname' '$hmdesc' '$inputmatrix' $rowOrderMethod $rowDistanceMeasure $rowAgglomerationMethod $columnOrderMethod $columnDistanceMeasure $columnAgglomerationMethod '$summarymethod' '$GALAXY_ROOT_DIR/tools/MDA_Heatmaps/' -->
+<command>./heatmap.sh  '$hmname' '$hmdesc' '$inputmatrix' $rowOrderMethod $rowDistanceMeasure $rowAgglomerationMethod $columnOrderMethod $columnDistanceMeasure $columnAgglomerationMethod '$summarymethod' './MDA_Heatmaps/'
+    #for $op in $operations
+       ${op.class_name}
+       ${op.repeatinput.file_name}
+       ${op.cat}
+      #end for
+ 	'$output' 
+ </command>
+	<stdio>
+      <exit_code range="1:" level="fatal" />
+	</stdio>
+  <inputs>
+    <param name="inputmatrix" type="data" label="Input Matrix" />
+    <param name="hmname" size="20" type="text" value="HM_name" label="User Defined HeatMap Name"/>
+    <param name="hmdesc" size="100" optional="true" type="text" value="_" label="Optional HeatMap Description"/>
+    <param name="summarymethod" 	type="select"  label="Data Summarization Method">
+		<option value="average">average</option>
+		<option value="sample">sample</option>
+		<option value="mode">mode</option>
+    </param>
+	<param name="rowOrderMethod" type="select" label="Row ordering method" help="Choices -- Hierarchical Clustering, Original Order, Random">
+		<option value="Hierarchical">Hierarchical Clustering</option>
+		<option value="Original">Original Order</option>	    
+		<option value="Random">Random</option>	    
+	</param>
+	<param name="rowDistanceMeasure" type="select"  label="Row Distance Metric" help="euclidean, binary, manhattan, maximum, canberra, minkowski, or correlation">
+		<option value="euclidean">Euclidean</option>
+		<option value="binary">Binary</option>
+		<option value="manhattan">Manhattan</option>
+		<option value="maximum">Maximum</option>
+		<option value="canberra">Canberra</option>	    
+		<option value="minkowski">Minkowski</option>	    
+		<option value="correlation">Correlation</option>	    
+	</param>
+	<param name="rowAgglomerationMethod" type="select"  label="Row Clustering Method" help="Choices: 'average' for Average Linkage, 'complete' for Complete Linkage, 'single' for Single Linkage, 'ward', 'mcquitty', 'median', or 'centroid'.">
+		<option value="average">Average Linkage</option>
+		<option value="complete">Complete Linkage</option>
+		<option value="single">Single Linkage</option>
+		<option value="ward">Ward</option>
+		<option value="mcquitty">Mcquitty</option>	    
+		<option value="median">Median</option>	    
+		<option value="centroid">Centroid</option>	    
+	</param>
+	<param name="columnOrderMethod" type="select" label="Column ordering method" help="Choices -- Hierarchical Clustering, Original Order, Random">
+		<option value="Hierarchical">Hierarchical Clustering</option>
+		<option value="Original">Original Order</option>	    
+		<option value="Random">Random</option>	    
+	</param>
+	<param name="columnDistanceMeasure" type="select"  label="Column Distance Metric" help="euclidean, binary, manhattan, maximum, canberra, minkowski, or correlation">
+		<option value="euclidean">Euclidean</option>
+		<option value="binary">Binary</option>
+		<option value="manhattan">Manhattan</option>
+		<option value="maximum">Maximum</option>
+		<option value="canberra">Canberra</option>	    
+		<option value="minkowski">Minkowski</option>	    
+		<option value="correlation">Correlation</option>	    
+	</param>
+	<param name="columnAgglomerationMethod" type="select"  label="Column Clustering Method" help="Choices: 'average' for Average Linkage, 'complete' for Complete Linkage, 'single' for Single Linkage, 'ward', 'mcquitty', 'median', or 'centroid'.">
+		<option value="average">Average Linkage</option>
+		<option value="complete">Complete Linkage</option>
+		<option value="single">Single Linkage</option>
+		<option value="ward">Ward</option>
+		<option value="mcquitty">Mcquitty</option>	    
+		<option value="median">Median</option>	    
+		<option value="centroid">Centroid</option>	    
+	</param>
+    <repeat name="operations" title="Covariate Bars">
+        <param name="class_name" size="20" type="text" value="" label="Axis Covariate Name"/>
+        <param name="repeatinput" type="data" format="text" label="Axis Covariate File"/>        
+		<param name="cat" type="select" label="Axis Covariate Type">
+		  <option value="row_categorical" >row categorical</option>
+		  <option value="row_continuous" >row continuous</option>
+		  <option value="column_categorical" >column categorical</option>
+		  <option value="column_continuous" >column continuous</option>
+		</param>
+    </repeat>       
+  </inputs>
+  <outputs>
+    <data name="output" label='${hmname}' format="zip"/>
+  </outputs>
+</tool>