Mercurial > repos > recetox > recetox_xmsannotator_simple

changeset 2:c72233720bdc draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/RECETOX/galaxytools/tree/master/tools/recetox-xmsannotator commit e547915248546657de13b91f28e5f4488ae8f7e0"
author recetox
date Thu, 12 Aug 2021 11:12:32 +0000
parents fa85bd6fd713
children b8b1b865c82d
files macros.xml recetox_xmsannotator_macros.xml recetox_xmsannotator_simple.xml tool-data/recetox_metabolite_databases.loc.sample tool_data_table_conf.xml.sample
diffstat 5 files changed, 106 insertions(+), 79 deletions(-) [+]
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/macros.xml	Thu Aug 12 11:12:32 2021 +0000
@@ -0,0 +1,57 @@
+<macros>
+    <token name="@TOOL_VERSION@">0.9.0</token>
+    <xml name="requirements">
+        <requirements>
+            <requirement type="package" version="0.9.0">r-recetox-xmsannotator</requirement>
+        </requirements>
+    </xml>
+
+    <xml name="inputs">
+        <param name="peak_table" type="data" format="csv,h5,parquet">
+            <label>Peak table</label>
+            <help><![CDATA[
+                A peak-intensity table such as outputted from apLCMS.
+                The file is required to contain the fields <em>mz</em> and <em>rt</em>.
+            ]]></help>
+        </param>
+        <param name="compound_table" type="data" format="csv,parquet">
+            <label>Compound database</label>
+            <help><![CDATA[
+                Database of compounds according to which the annotation is performed.
+                The database is required to contain the fields <em>compound_id</em>, <em>monoisotopic_mass</em>, and <em>molecular_formula</em>.
+            ]]></help>
+        </param>
+        <param name="adduct_table" type="data" format="csv,parquet" optional="true">
+            <label>Adduct database (optional)</label>
+            <help><![CDATA[
+                Database of adduct which is combined with the database of compound to form a molecule-adduct pairs.
+                The database is required to contain <em>adduct</em>, <em>charge</em>, <em>mass</em>, and <em>n_molecules</em>.
+            ]]></help>
+        </param>
+    </xml>
+
+    <xml name="outputs">
+        <data name="annotation_parquet" format="parquet"/>
+    </xml>
+
+    <xml name="tolerance">
+        <param name="mass_tolerance_ppm" type="integer" min="0" value="5">
+            <label>Mass tolerance [ppm]</label>
+            <help>Mass tolerance in ppm for database matching.</help>
+        </param>
+        <yield/>
+    </xml>
+    <token name="@HELP@">
+        <![CDATA[
+        Annotate the peak intensity table (e.g. from an apLCMS run) with compounds from the compounds database
+        using advanced methods.
+
+        The annotation process generates all possible compound-adduct pairs and matches those pairs to the measured
+        peaks. A compound-adduct pair is pronounced as a match to a certain peak when the difference of their masses are
+        withing some tolerance.
+        ]]>
+    </token>
+    <xml name="citations">
+        <citation type="doi">10.1021/acs.analchem.6b01214</citation>
+    </xml>
+</macros>
--- a/recetox_xmsannotator_macros.xml	Thu Sep 17 08:39:26 2020 +0000
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,36 +0,0 @@
-<macros>
-    <xml name="requirements">
-        <requirements>
-            <requirement type="package">recetox_datatypes</requirement>
-            <container type="docker">registry.gitlab.ics.muni.cz:443/recetox/mass-spectrometry/xmsannotator:v2.0</container>
-        </requirements>
-    </xml>
-
-    <xml name="peaks">
-        <param name="peaks" type="data" format="peak_table.feather" label="Peak intensity table"/>
-    </xml>
-
-    <xml name="metabolites">
-        <!-- <param name="metabolites" type="select" label="Metabolite database">
-            <options from_data_table="recetox_metabolite_databases">
-                <column name="name" index="2"/>
-                <column name="value" index="3"/>
-            </options>
-        </param> -->
-        <param name="metabolites" type="data" format="metabolites.h5" label="Metabolite database"/>
-    </xml>
-
-    <xml name="mz_tolerance">
-        <param name="mz_tolerance" type="float" value="10" min="0" label="Mass tolerance [ppm]" help="Mass tolerance in ppm for database matching."/>
-    </xml>
-
-    <xml name="annotation_output">
-        <data format="annotated_peak_list.feather" name="annotation"/>
-    </xml>
-
-    <xml name="citations">
-        <citations>
-            <citation type="doi">10.1021/acs.analchem.6b01214</citation>
-        </citations>
-    </xml>
-</macros>
\ No newline at end of file
--- a/recetox_xmsannotator_simple.xml	Thu Sep 17 08:39:26 2020 +0000
+++ b/recetox_xmsannotator_simple.xml	Thu Aug 12 11:12:32 2021 +0000
@@ -1,32 +1,49 @@
-<tool id="recetox_xmsannotator_simple" name="xmsannotator - simple" version="v2.0">
-    <macros>
-        <import>recetox_xmsannotator_macros.xml</import>
-    </macros>
-
-    <expand macro="requirements"/>
-
-    <command detect_errors="aggressive"><![CDATA[
-        Rscript -e 'annotation <- xmsannotator::simple_annotation(
-                        peaks = arrow::read_feather("$peaks"),
-                        adducts = rhdf5::h5read("$adducts", "adducts"),
-                        metabolites = rhdf5::h5read("$metabolites", "metabolites"),
-                        mz_tolerance = as.double($mass_tolerance)
-                   )'
-                -e 'arrow::write_feather(annotation, "$annotation")'
-    ]]></command>
-
-    <inputs>
-        <param name="peaks" type="data" format="feather" label="Peaks" help="Table of peak intensities."/>
-        <param name="metabolites" type="data" format="h5" label="Metabolite database"/>
-        <param name="adducts" type="data" format="h5" label="Adduct database"/>
-
-        <param name="mass_tolerance" type="float" value="10" min="0" label="Mass tolerance [ppm]" help="Mass tolerance in ppm for database matching."/>
-    </inputs>
-
-    <outputs>
-        <data format="feather" name="annotation"/>
-    </outputs>
-
-    <help><![CDATA[]]></help>
-    <expand macro="citations"/>
-</tool>
\ No newline at end of file
+<tool id="recetox_xmsannotator_simple" name="RECETOX xMSannotator simple" version="@TOOL_VERSION@+galaxy0">
+    <description>annotate peak intensity table</description>
+    <macros>
+        <import>macros.xml</import>
+    </macros>
+    <expand macro="requirements" />
+
+    <command detect_errors="aggressive"><![CDATA[
+        Rscript $wrapper
+    ]]></command>
+
+    <configfiles>
+        <configfile name="wrapper"><![CDATA[
+            library(xmsannotator)
+
+            annotation <- simple_annotation(
+            #if $peak_table.is_of_type("h5")
+                peak_table = load_peak_table_hdf("${peak_table}"),
+            #elif $peak_table.is_of_type("parquet")
+                peak_table = load_peak_table_parquet("${peak_table}"),
+            #end if
+                adduct_table = load_adduct_table_parquet("${adduct_table}"),
+                compound_table = load_compound_table_parquet("${compound_table}"),
+                mass_tolerance = 1e-6 * ${mass_tolerance_ppm}
+            )
+
+            save_parquet(data = annotation, file = "${annotation_parquet}")
+        ]]></configfile>
+    </configfiles>
+
+    <inputs>
+        <expand macro="inputs"/>
+        <expand macro="tolerance"/>
+    </inputs>
+
+    <outputs>
+        <expand macro="outputs"/>
+    </outputs>
+
+    <help>        
+        <![CDATA[
+        @HELP@
+        ]]>
+    </help>
+
+    <citations>
+        <expand macro="citations"/>
+    </citations>
+</tool>
--- a/tool-data/recetox_metabolite_databases.loc.sample	Thu Sep 17 08:39:26 2020 +0000
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,5 +0,0 @@
-# This file has the format (white space characters are
-# TAB characters):
-#
-# <value>	<name>	<date>	<path>
-#
\ No newline at end of file
--- a/tool_data_table_conf.xml.sample	Thu Sep 17 08:39:26 2020 +0000
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,6 +0,0 @@
-<tables>
-    <table name="recetox_metabolite_databases" comment_char="#">
-        <columns>value, date, name, path</columns>
-        <file path="tool-data/recetox_metabolite_databases.loc" />
-    </table>
-</tables>
\ No newline at end of file