# HG changeset patch
# User artbio
# Date 1549841511 18000
# Node ID 008de522b3eaf66d61b37f40b4da1c2f7a2b5de4
# Parent  243ed53cbc0d860af93ab34ae6f481f223fe164e
planemo upload for repository https://github.com/ARTbio/tools-artbio/tree/master/tools/sr_bowtie_dataset_annotation commit 53e9bab2c20411f34ac09688de2f2cc8ae8c46a4

diff -r 243ed53cbc0d -r 008de522b3ea barplot.r
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/barplot.r	Sun Feb 10 18:31:51 2019 -0500
@@ -0,0 +1,48 @@
+if (length(commandArgs(TRUE)) == 0) {
+  system("Rscript barplot.r -h", intern = F)
+  q("no")
+}
+
+
+# load packages that are provided in the conda env
+options( show.error.messages=F,
+       error = function () { cat( geterrmessage(), file=stderr() ); q( "no", 1, F ) } )
+loc <- Sys.setlocale("LC_MESSAGES", "en_US.UTF-8")
+warnings()
+library(optparse)
+library(ggplot2)
+library(scales)
+
+
+
+#Arguments
+option_list = list(
+  make_option(
+    c("-i", "--input"),
+    default = NA,
+    type = 'character',
+    help = "Input file that contains count data (no header)"
+  ),
+  make_option(
+    c("-o", "--barplot"),
+    default = NA,
+    type = 'character',
+    help = "PDF output file"
+  )
+)
+
+opt = parse_args(OptionParser(option_list = option_list),
+                 args = commandArgs(trailingOnly = TRUE))
+
+
+## 
+annotations = read.delim(opt$input, header=F)
+colnames(annotations) = c("class", "counts")
+annotations = cbind(annotations, fraction=annotations$counts/annotations$counts[1])
+annotations = annotations[-1,]
+# ggplot2 plotting
+ggplot(annotations, aes(x="classes", y=fraction, fill=class)) +
+geom_bar(width = .7, position=position_stack(), stat = "identity") +
+geom_text(aes(label = percent(fraction)), position = position_stack(vjust = 0.5),size = 4)
+ggtitle('Class proportions') 
+ggsave(file=opt$barplot, device="pdf")
diff -r 243ed53cbc0d -r 008de522b3ea sr_bowtie_dataset_annotation.xml
--- a/sr_bowtie_dataset_annotation.xml	Sun Feb 10 14:07:12 2019 -0500
+++ b/sr_bowtie_dataset_annotation.xml	Sun Feb 10 18:31:51 2019 -0500
@@ -1,7 +1,10 @@
-<tool id="sr_bowtie_dataset_annotation" name="Annotate smRNA dataset" version="2.0.2">
+<tool id="sr_bowtie_dataset_annotation" name="Annotate smRNA dataset" version="2.1.0">
   <description>by iterative alignments with sRbowtie</description>
   <requirements>
         <requirement type="package" version="1.1.2">bowtie</requirement>
+        <requirement type="package" version="1.3.2">r-optparse</requirement>
+        <requirement type="package" version="2.2.1">r-ggplot2</requirement>
+        <requirement type="package" version="0.4.1">r-scales</requirement>
   </requirements>
   <command  detect_errors="exit_code"><![CDATA[
         #if $refGenomeSource1.genomeSource == "history":
@@ -11,9 +14,9 @@
         #else:
             #set index_path = $refGenomeSource1.index.fields.path
         #end if
-        #if $input.extension == "fasta":
+        #if $input.is_of_type('fasta'):
             #set format = "-f"
-        #elif $input.extension == "fastq":
+        #elif $input.is_of_type('fastq'):
             #set format = "-q"
         #end if
         #if $format == '-f':
@@ -60,7 +63,8 @@
             echo -e "$i.ownFile.name\t\${class_aligned}\n" >> $output &&
         #end for
         remaining=\$(( \$(wc -l < class_unmatched.fa)/2)) &&
-        echo -e "Unmatched to previous indexes\t\${remaining}\n" >> $output
+        echo -e "Not classified\t\${remaining}\n" >> $output &&
+        Rscript $__tool_directory__/barplot.r --input $output --barplot $barplot
         ]]></command>
   <inputs>
       <param name="input" type="data" format="fasta,fastq" label="Input file: reads clipped from their adapter" help="Only with clipped, raw fasta or fastq files"/>
@@ -93,12 +97,13 @@
 <!-- End of other bowtie index selections -->
    </inputs>
    <outputs>
-   <data format="tabular" name="output" label="Cascade Annotation Analysis">
-       <actions>
-           <action name="column_names" type="metadata" default="Reference Index,Number of reads" />
-       </actions>
-    </data>
-   </outputs>
+       <data format="tabular" name="output" label="Cascade Annotation Analysis">
+           <actions>
+               <action name="column_names" type="metadata" default="Reference Index,Number of reads" />
+           </actions>
+        </data>
+        <data name="barplot" format="pdf" label="barplot from ${on_string}" />
+    </outputs>
     <tests>
         <test>
             <param name="input" value ="sample1.fa" ftype="fasta" />
@@ -107,6 +112,7 @@
             <param name="AdditionalQueries_0|ownFile" value="dme_miR21_hairpin.fa" ftype="fasta" />
             <param name="AdditionalQueries_1|ownFile" value="Ensembl_transposon_set.fa" ftype="fasta" />
             <output name="output" ftype="tabular" file="sample1_output.tab" />
+            <output name="barplot" ftype="pdf" file="sample1_output.pdf" />
         </test>
         <test>
             <param name="input" value ="sample.fastq" ftype="fastq" />
@@ -115,6 +121,7 @@
             <param name="AdditionalQueries_0|ownFile" value="dme_miR21_hairpin.fa" ftype="fasta" />
             <param name="AdditionalQueries_1|ownFile" value="Ensembl_transposon_set.fa" ftype="fasta" />
             <output name="output" ftype="tabular" file="sample_output.tab" />
+            <output name="barplot" ftype="pdf" file="sample_output.pdf" />
         </test>
     </tests>
   <help>
diff -r 243ed53cbc0d -r 008de522b3ea test-data/sample1_output.pdf
Binary file test-data/sample1_output.pdf has changed
diff -r 243ed53cbc0d -r 008de522b3ea test-data/sample1_output.tab
--- a/test-data/sample1_output.tab	Sun Feb 10 14:07:12 2019 -0500
+++ b/test-data/sample1_output.tab	Sun Feb 10 18:31:51 2019 -0500
@@ -4,5 +4,5 @@
 
 Ensembl_transposon_set.fa	2689
 
-Unmatched to previous indexes	265
+Not classified	265
 
diff -r 243ed53cbc0d -r 008de522b3ea test-data/sample_output.pdf
Binary file test-data/sample_output.pdf has changed
diff -r 243ed53cbc0d -r 008de522b3ea test-data/sample_output.tab
--- a/test-data/sample_output.tab	Sun Feb 10 14:07:12 2019 -0500
+++ b/test-data/sample_output.tab	Sun Feb 10 18:31:51 2019 -0500
@@ -4,5 +4,5 @@
 
 Ensembl_transposon_set.fa	560
 
-Unmatched to previous indexes	78
+Not classified	78