Mercurial > repos > iuc > medaka_variant

changeset 4:2308bc523659 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/medaka commit c02f74da66f51b431b7cd8bd0d1cffde267e3332"
author iuc
date Tue, 13 Oct 2020 16:00:16 +0000
parents e86fcef8ed91
children 19f3b583a9b3
files macros.xml variant.xml
diffstat 2 files changed, 103 insertions(+), 26 deletions(-) [+]
line wrap: on
line diff
--- a/macros.xml	Tue Sep 01 03:08:24 2020 -0400
+++ b/macros.xml	Tue Oct 13 16:00:16 2020 +0000
@@ -5,7 +5,7 @@
     <xml name="requirements">
         <requirements>
             <requirement type="package" version="@TOOL_VERSION@">medaka</requirement>
-            <requirement type="package" version="1.4.1">scipy</requirement>
+            <yield />
         </requirements>
     </xml>
     <xml name="version_command">
--- a/variant.xml	Tue Sep 01 03:08:24 2020 -0400
+++ b/variant.xml	Tue Oct 13 16:00:16 2020 +0000
@@ -1,40 +1,75 @@
 <?xml version="1.0"?>
-<tool id="medaka_variant" name="medaka: Variant" version="@TOOL_VERSION@+galaxy1" profile="@PROFILE@">
+<tool id="medaka_variant" name="medaka variant tool" version="@TOOL_VERSION@+galaxy2" profile="@PROFILE@">
     <description>Probability decoding</description>
     <macros>
         <import>macros.xml</import>
     </macros>
-    <expand macro="requirements"/>
+    <expand macro="requirements">
+        <requirement type="package" version="1.4.1">scipy</requirement>
+    </expand>
+
     <expand macro="version_command"/>
     <command detect_errors="exit_code"><![CDATA[
 ## initialize
 @REF_FASTA@
 
-## run
-medaka variant
-## optional
---debug
-#if $region
-    --region '${region}'
+
+#if $pool.pool_mode == "Yes":
+    ## run
+    medaka variant
+    ## optional
+    --debug
+    #if $regions
+        --regions '${regions}'
+    #end if
+    $verbose
+    ## required
+    reference.fa
+    #for $current in $pool.inputs
+        '$current'
+    #end for
+    '$out_result' ## output
+
+    2>&1 | tee '$out_log'
+#elif $pool.pool_mode == "No":
+    ## run
+    medaka variant
+    ## optional
+    --debug
+    #if $regions
+        --regions '${regions}'
+    #end if
+    $verbose
+    ## required
+    reference.fa
+    '$pool.input'
+    '$out_result' ##output
+    2>&1 | tee '$out_log'
 #end if
-$verbose
-## required
-reference.fa
-#for $current in $inputs
-    '$current'
-#end for
-'$out_result' ## output
-
-2>&1 | tee '$out_log'
 
 #if $out_annotated
-    && python '$__tool_directory__/annotateVCF.py' '$out_result' '$in_mpileup' '$out_annotated'
+    && samtools faidx reference.fa
+    && grep -v "^#" '$out_result' > tmp.txt
+    && cut -f 1,2 tmp.txt > positions.txt
+    && samtools mpileup -d 0 --positions positions.txt -Q $output_annotated.min_basequal -q $output_annotated.min_mapqual -f reference.fa '$output_annotated.in_bam' > mpileup.txt
+    && python '$__tool_directory__/annotateVCF.py' '$out_result' mpileup.txt '$out_annotated'
 #end if
     ]]></command>
     <inputs>
-        <param argument="inputs" type="data" format="h5" multiple="true" label="Select consensus file(s)"/>
+        <conditional name="pool">
+            <param name="pool_mode"  type="select" label="Are you pooling HDF5 datasets?">
+                <option value="No" selected="True">No</option>
+                <option value="Yes">Yes</option>
+            </param>
+            <when value="Yes">
+                <param name="inputs" type="data" format="h5" multiple="true" label="Select consensus file(s)"/>
+            </when>
+            <when value="No">
+                <param name="input" type="data" format="h5" label="Select consensus file(s)"/>
+            </when>
+        </conditional>
         <expand macro="reference"/>
-        <param argument="--region" type="text" value="" optional="true" label="Set reference names to limit variant calling" help="Separated by ','.">
+        <param argument="--regions" type="text" value="" optional="true" label="Set reference names to limit variant calling" help="Separated by ','.">
             <sanitizer invalid_char="">
                 <valid initial="string.ascii_letters,string.digits">
                     <add value="_"/>
@@ -44,15 +79,17 @@
             </sanitizer>
         </param>
         <param argument="--verbose" type="boolean" truevalue="--verbose" falsevalue="" label="Populate VCF info fields?"/>
-
         <conditional name="output_annotated">
-            <param name="output_annotated_select" type="select" label="Output annotated VCF?" help="Annotate allele frequency, depth of coverage, etc for each variant (requires pileup file)">
+            <param name="output_annotated_select" type="select" label="Output annotated VCF?" help="Annotate allele frequency, depth of coverage, etc for each variant (requires BAM file)">
                 <option value="true" selected="true">Output annotated VCF</option>
                 <option value="false">Don't output annotated VCF</option>
             </param>
             <when value="true">
-                <param name="in_mpileup" type = "data" format="pileup" optional="false" label="Mpileup to annotate the VCF"/>
+                <param name="in_bam" type="data" format="bam" optional="false" label="BAM to annotate the VCF" />
+                <param name="min_basequal" type="integer" value="0" min="0" max="50" label="Minimum base quality" help="The minimum base quality (default: 0) at a given variant position required to include a read in the calculation of variant statistics (samtools mpileup -Q)" />
+                <param name="min_mapqual" type="integer" value="0" min="0" max="60" label="Minimum mapping quality" help="The minimum mapping quality (default: 0) required for a read to be considered in calculation of variant statistics (samtools mpileup -q)" />
             </when>
+            <when value="false"/>
         </conditional>
         <param name="output_log_bool" type="boolean" label="Output log file?" checked="true"/>
     </inputs>
@@ -70,14 +107,54 @@
     <tests>
         <!-- #1 default -->
         <test expect_num_outputs="3">
-            <param name="inputs" value="medaka_test.hdf"/>
+            <conditional name="pool">
+                <param name="pool_mode" value="Yes"/>
+                <param name="inputs" value="medaka_test.hdf,medaka_test.hdf"/>
+            </conditional>
             <conditional name="reference_source">
                 <param name="reference_source_selector" value="history"/>
                 <param name="ref_file" value="ref.fasta"/>
             </conditional>
             <conditional name="output_annotated">
                 <param name="output_annotated_select" value="true"/>
-                <param name="in_mpileup" value="annotate_vcf_test.pileup"/>
+                <param name="in_bam" value="medaka_test.bam"/>
+            </conditional>
+            <param name="output_log_bool" value="true"/>
+            
+            <output name="out_result">
+                <assert_contents>
+                    <has_n_lines n="9"/>
+                    <has_line line="##fileformat=VCFv4.1" />
+                    <has_line line="##medaka_version=1.0.3" />
+                </assert_contents>
+            </output>
+            <output name="out_annotated">
+                <assert_contents>
+                    <has_n_lines n="16"/>
+                    <has_line line="##fileformat=VCFv4.1" />
+                    <has_line line="##medaka_version=1.0.3" />
+                    <has_line line="##annotateVCFVersion=0.1" />
+                </assert_contents>
+            </output>
+            <output name="out_log">
+                <assert_contents>
+                    <has_n_lines n="10" />
+                </assert_contents>
+            </output>
+        </test>
+        <!--No pooling-->
+        <test expect_num_outputs="3">
+            <conditional name="pool">
+                <param name="pool_mode" value="No"/>
+                <param name="input" value="medaka_test.hdf"/>
+            </conditional>
+            <conditional name="reference_source">
+                <param name="reference_source_selector" value="history"/>
+                <param name="ref_file" value="ref.fasta"/>
+            </conditional>
+            <conditional name="output_annotated">
+                <param name="output_annotated_select" value="true"/>
+                <param name="in_bam" value="medaka_test.bam"/>
             </conditional>
             <param name="output_log_bool" value="true"/>