Mercurial > repos > iuc > biobox_add_taxid

changeset 2:450a61fc097f 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/amber/ commit 956c476d4eeed0334582438d715da88e170212c6
author iuc
date Sun, 08 Sep 2024 14:08:45 +0000
parents d368e312aebd
children 8d0b2def5e65
files biobox_add_taxid.xml test-data/test_taxonkit_file.tsv test-data/test_taxonkit_file_1.tsv test-data/test_taxonkit_file_2.tsv
diffstat 4 files changed, 41 insertions(+), 32 deletions(-) [+]
line wrap: on
line diff
--- a/biobox_add_taxid.xml	Fri Aug 30 14:39:04 2024 +0000
+++ b/biobox_add_taxid.xml	Sun Sep 08 14:08:45 2024 +0000
@@ -2,7 +2,7 @@
     <description>Add taxid output from BAT or GTDB to biobox binning data</description>
     <macros>
         <import>macros.xml</import>
-        <token name="@SCRIPT_VERSION@">0.4</token>
+        <token name="@SCRIPT_VERSION@">0.6</token>
     </macros>
     <requirements>
         <requirement type="package" version="@SCRIPT_VERSION@">biobox_add_taxid</requirement>
@@ -10,7 +10,7 @@
     <command detect_errors="exit_code">
     <![CDATA[
 
-        mkdir -p input &&
+        mkdir -p input taxonkit gtdb_to_taxdump &&
 
         ln -s '${biobox_file}' '$biobox_file.element_identifier' &&
 
@@ -18,13 +18,22 @@
             ln -s '$f' 'input/${i}.tsv' &&
         #end for
 
+        #if $tool_type.is_select == 'GTDB':
+            #for $i, $f in enumerate($gtdb_to_taxdump):
+                ln -s '$f' 'gtdb_to_taxdump/gtdb_to_taxdump_${i}.tsv' &&
+            #end for
+            #for $i, $f in enumerate($taxonkit):
+                ln -s '$f' 'taxonkit/taxonkit_${i}.tsv' &&
+            #end for
+        #end if
+
         biobox_add_taxid.py
         '$biobox_file.element_identifier'
         '$tool_type.is_select'
         'input'
         #if $tool_type.is_select == 'GTDB':
-            -g '$gtdb_to_taxdump'
-            -t '$taxonkit'
+            -g 'gtdb_to_taxdump'
+            -t 'taxonkit'
             -c $column
         #end if
         &&
@@ -45,8 +54,8 @@
                 <option value="GTDB">GTDB</option>
             </param>
             <when value="GTDB">
-                <param argument="--gtdb_to_taxdump" type="data" format="tabular" label="Input the output from gtdb_to_taxdump here" help="Use the output from gtdb_to_taxdump here since we need the mapped names from GTDB to NCBI to get the arcoding taxids from NCBI"/>
-                <param argument="--taxonkit" type="data" format="tabular" label="Input the output from Taxonkit here" help="Use the output from Taxonkit here since the need the mapped NCBI names to the arcording taxids"/>
+                <param argument="--gtdb_to_taxdump" type="data" multiple="true" format="tabular" label="Input the output from gtdb_to_taxdump here" help="Use the output from gtdb_to_taxdump here since we need the mapped names from GTDB to NCBI to get the arcoding taxids from NCBI"/>
+                <param argument="--taxonkit" type="data" format="tabular" multiple="true" label="Input the output from Taxonkit here" help="Use the output from Taxonkit here since the need the mapped NCBI names to the arcording taxids"/>
                 <param argument="--column" type="data_column" data_ref="taxonkit" label="Input the colum with the NCBI names"/>
                 <param name="input" type="data" multiple="true" format="tabular" label="Input the GTDB-Tk summary file(s) here"
                     help="Use the GTDB-Tk file(s) which are coresponding with the binning file!"/>
@@ -74,7 +83,7 @@
             <conditional name="tool_type">
                 <param name="is_select" value="GTDB"/>
                 <param name="gtdb_to_taxdump" ftype="tabular" value="test_gtdb_to_taxdump_file.tsv"/>
-                <param name="taxonkit" ftype="tabular" value="test_taxonkit_file.tsv"/>
+                <param name="taxonkit" ftype="tabular" value="test_taxonkit_file_1.tsv,test_taxonkit_file_2.tsv"/>
                 <param name="column" value="1"/>
                 <param name="input" ftype="tabular" value="test_gtdb_file.tsv"/>
             </conditional>
--- a/test-data/test_taxonkit_file.tsv	Fri Aug 30 14:39:04 2024 +0000
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,25 +0,0 @@
-unclassified	12908
-Xylella fastidiosa	2371
-Paramaledivibacter caminithermalis	191027
-Flavisolibacter ginsengisoli	462367
-Bacilli	91061
-unclassified	12908
-unclassified	12908
-Tepidibacter formicigenes	227138
-unclassified	12908
-unclassified	12908
-Andreprevotia lacus	1121000
-Nonlabens	363408
-Pseudomonas aeruginosa	287
-Paracoccus	265
-Paracoccus	249411
-unclassified	12908
-Anaerobranca californiensis	182411
-Hydrotalea sandarakina	1004304
-Selenomonadaceae	1843491
-unclassified	12908
-Lampropedia hyalina	198706
-unclassified	12908
-Anaerosporobacter mobilis	264463
-Meinhardsimonia xiamenensis	990712
-Desulfatibacillum alkenivorans	259354
\ No newline at end of file
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/test-data/test_taxonkit_file_1.tsv	Sun Sep 08 14:08:45 2024 +0000
@@ -0,0 +1,11 @@
+unclassified	12908
+Xylella fastidiosa	2371
+Paramaledivibacter caminithermalis	191027
+Flavisolibacter ginsengisoli	462367
+Bacilli	91061
+unclassified	12908
+unclassified	12908
+Tepidibacter formicigenes	227138
+unclassified	12908
+unclassified	12908
+Andreprevotia lacus	1121000
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/test-data/test_taxonkit_file_2.tsv	Sun Sep 08 14:08:45 2024 +0000
@@ -0,0 +1,14 @@
+Nonlabens	363408
+Pseudomonas aeruginosa	287
+Paracoccus	265
+Paracoccus	249411
+unclassified	12908
+Anaerobranca californiensis	182411
+Hydrotalea sandarakina	1004304
+Selenomonadaceae	1843491
+unclassified	12908
+Lampropedia hyalina	198706
+unclassified	12908
+Anaerosporobacter mobilis	264463
+Meinhardsimonia xiamenensis	990712
+Desulfatibacillum alkenivorans	259354