diff pangolin.xml @ 20:14ae456b8cc5 draft

"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/pangolin commit 21834c24b94f942ed759bd7a2fcf0b3b4b5fd839"

--- a/pangolin.xml	Thu Apr 21 11:40:56 2022 +0000
+++ b/pangolin.xml	Tue May 03 19:27:06 2022 +0000
@@ -1,4 +1,4 @@
-<tool id="pangolin" name="Pangolin" version="@TOOL_VERSION@+galaxy0" profile="20.01">
+<tool id="pangolin" name="Pangolin" version="@TOOL_VERSION@+galaxy1" profile="20.01">
     <description>Phylogenetic Assignment of Outbreak Lineages</description>
     <macros>
         <token name="@TOOL_VERSION@">4.0.5</token>
@@ -10,6 +10,11 @@
     </requirements>
     <version_command><![CDATA[pangolin --version]]></version_command>
     <command detect_errors="exit_code"><![CDATA[
+      #if $engine.use_assignment_cache and str($db.source) != "download":
+        ## This is no good. Better to fail immediately instead of downloading a lot of data first.
+        echo "Using the latest assignment cache requires downloading the latest version of pangolin-data." 1>&2; exit 1
+      #else:
+        ## Sanity chceck was ok, lets do the real thing ...
         #if str($db.source) == "download"
             ## Pangolin version 4 tries to update from an existing directory
             mkdir datadir &&
@@ -17,6 +22,21 @@
         #else if str($db.source) == "builtin"
             ln -s $db.db_release.fields.path datadir &&
         #end if
+        #if $engine.use_assignment_cache:
+            ## We need to install also the latest UShER assignment cache data.
+            ## Pangolin has functionality to do so, but uses it incorrectly.
+            ## We use the pangolin function to install into --datadir here,
+            ## then point pangolin to the downloaded file later using
+            ## its --assignment-cache parameter
+
+            ## Create a "honeypot" package that will be picked up by pangolin,
+            ## but will trigger a download because of missing __version__ info.
+            mkdir pangolin_assignment &&
+            touch pangolin_assignment/__init__.py &&
+            ## Call pangolin's assignment cache install function, but
+            ## override pip's install path
+            PIP_TARGET="datadir" PIP_UPGRADE=1 python -c "from pangolin.utils import update; update.install_pangolin_assignment()" &&
+        #end if
         pangolin
         --threads \${GALAXY_SLOTS:-1}
         --tempdir "\${TMPDIR:-.}"
@@ -26,20 +46,25 @@
         --analysis-mode $engine.analysis_mode
         #if str($engine.analysis_mode) == 'usher':
             $engine.use_assignment_cache
+            #if $engine.use_assignment_cache:
+                ## Point pangolin to the assignment cache file we've downloaded before
+                --assignment-cache datadir/pangolin_assignment/usher_assignments.cache.csv.gz
+            #end if
         #end if
         #if $alignment:
             $alignment --alignment-file '$align1'
         #end if
-        --outfile report.csv 
+        --outfile report.csv
         --max-ambig $max_ambig
         --min-length $min_length
         $expanded_lineage
         '$input1'
         && csvtk csv2tab report.csv
         #if not $include_header:
-            | tail -n+2 
+            | tail -n+2
         #end if
         > '$output1'
+      #end if
     ]]></command>
     <inputs>
         <param type="data" name="input1" format="fasta" label="Input FASTA File(s)" />
@@ -50,8 +75,8 @@
                 <option value="pangolearn">pangoLEARN</option>
             </param>
             <when value="usher">
-                <param argument="--use-assignment-cache" type="boolean" truevalue="--add-assignment-cache --use-assignment-cache" falsevalue="" label="Use latest UShER assignment cache"
-                help="Get the latest UShER assignment cache from the pangolin-assignment online repository and use it to speed up UShER lineage assignment. Note: Downloading the cached assignments will only pay off for large numbers of input samples. Also note that using the latest assignment cache in combination with the built-in or a cached pangolin-data source (see option below), will make your otherwise reproducible results dependent on an external data source." />
+                <param argument="--use-assignment-cache" type="boolean" truevalue="--use-assignment-cache" falsevalue="" label="Use latest UShER assignment cache"
+                help="Get the latest UShER assignment cache from the pangolin-assignment online repository and use it to speed up UShER lineage assignment. Note: Downloading the cached assignments will only pay off for large numbers of input samples. Also note that using the latest assignment cache will require you to select the 'Download latest from web' option for the pangolin-data source below because assignment cache and pangolin-data need to be synchronized." />
             </when>
             <when value="pangolearn" />
         </conditional>
@@ -82,7 +107,7 @@
         </conditional>
         <param argument="--alignment" type="boolean" truevalue="--alignment" falsevalue="" label="Output multiple sequence alignment of input sequences" />
         <param argument="--max-ambig" type="float" value="0.3" min="0" max="1" label="Maximum proportion of Ns allowed" help="Maximum proportion of Ns allowed for pangolin to attempt assignment" />
-        <param argument="--min-length" type="integer" value="25000" min="0" max="29903" label="Minimum query length allowed" help="Minimum query length allowed for pangolin to attempt assignment. Please note that in the current implementation this parameter is used to calculate an alternate value for the 'Maximum proportion of Ns allowed' parameter as 1-(minlen/reflen). The smaller of the two will be used." />
+        <param argument="--min-length" type="integer" value="0" min="0" max="29903" label="Minimum query length allowed" help="Minimum query length allowed for pangolin to attempt assignment. Please note that in the current implementation this parameter is used to calculate an alternate value for the 'Maximum proportion of Ns allowed' parameter as 1-(minlen/reflen). The smaller of the two will be used." />
         <param argument="--expanded-lineage" type="boolean" truevalue="--expanded-lineage" falsevalue="" label="Add expanded lineage column to output" help="Optional expanded lineage information as defined in the alias.json file in pangolin-data can be appended as an additional column to the output." />
         <param name="include_header" type="boolean" truevalue="true" falsevalue="false" label="Include header line in output file" />
     </inputs>