Mercurial > repos > althonos > gecco

comparison gecco.xml @ 14:56b924f62165 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"Update tests files for Galaxy tool wrapper"
author althonos
date Tue, 05 Apr 2022 23:18:49 +0000
parents d64fe390f3c9
children cc91d730cc4f
comparison
equal deleted inserted replaced
13:fde43648cba0 14:56b924f62165
1 <?xml version='1.0' encoding='utf-8'?> 1 <?xml version='1.0' encoding='utf-8'?>
2 <tool id="gecco" name="GECCO" version="0.8.10" python_template_version="3.5"> 2 <tool id="gecco" name="GECCO" version="0.9.1" python_template_version="3.5">
3 <description>is a fast and scalable method for identifying putative novel Biosynthetic Gene Clusters (BGCs) in genomic and metagenomic data using Conditional Random Fields (CRFs).</description> 3 <description>is a fast and scalable method for identifying putative novel Biosynthetic Gene Clusters (BGCs) in genomic and metagenomic data using Conditional Random Fields (CRFs).</description>
4 <requirements> 4 <requirements>
5 <requirement type="package" version="0.8.10">gecco</requirement> 5 <requirement type="package" version="0.9.1">gecco</requirement>
6 </requirements> 6 </requirements>
7 <version_command>gecco --version</version_command> 7 <version_command>gecco --version</version_command>
8 <command detect_errors="aggressive"><![CDATA[ 8 <command detect_errors="aggressive"><![CDATA[
9 9
10 #if str($input.ext) == 'genbank': 10 #if str($input.ext) == 'genbank':
16 16
17 gecco -vv run 17 gecco -vv run
18 --format $input.ext 18 --format $input.ext
19 --genome input_tempfile.$file_extension 19 --genome input_tempfile.$file_extension
20 --postproc $postproc 20 --postproc $postproc
21 --edge-distance $edge_distance 21 --force-tsv
22 --force-clusters-tsv 22 #if $edge_distance
23 --edge-distance $edge_distance
24 #end if
23 #if $mask 25 #if $mask
24 --mask 26 --mask
25 #end if 27 #end if
26 #if $cds: 28 #if $cds:
27 --cds $cds 29 --cds $cds
31 #end if 33 #end if
32 #if $antismash_sideload: 34 #if $antismash_sideload:
33 --antismash-sideload 35 --antismash-sideload
34 #end if 36 #end if
35 37
38 && mv input_tempfile.genes.tsv '$genes'
36 && mv input_tempfile.features.tsv '$features' 39 && mv input_tempfile.features.tsv '$features'
37 && mv input_tempfile.clusters.tsv '$clusters' 40 && mv input_tempfile.clusters.tsv '$clusters'
38 #if $antismash_sideload 41 #if $antismash_sideload
39 && mv input_tempfile.sideload.json '$sideload' 42 && mv input_tempfile.sideload.json '$sideload'
40 #end if 43 #end if
47 <param argument="--threshold" type="float" min="0" max="1" value="" optional="true" label="Probability threshold for cluster detection"/> 50 <param argument="--threshold" type="float" min="0" max="1" value="" optional="true" label="Probability threshold for cluster detection"/>
48 <param argument="--postproc" type="select" label="Post-processing method for gene cluster validation"> 51 <param argument="--postproc" type="select" label="Post-processing method for gene cluster validation">
49 <option value="antismash">antiSMASH</option> 52 <option value="antismash">antiSMASH</option>
50 <option value="gecco" selected="true">GECCO</option> 53 <option value="gecco" selected="true">GECCO</option>
51 </param> 54 </param>
52 <param argument="--edge-distance" type="integer" min="0" value="10" label="Number of genes from the contig edges to filter out"/> 55 <param argument="--edge-distance" type="integer" min="0" optional="true" value="" label="Number of genes from the contig edges to filter out"/>
53 <param argument="--antismash-sideload" type="boolean" checked="false" label="Generate an antiSMASH v6 sideload JSON file"/> 56 <param argument="--antismash-sideload" type="boolean" checked="false" label="Generate an antiSMASH v6 sideload JSON file"/>
54 </inputs> 57 </inputs>
55 <outputs> 58 <outputs>
56 <collection name="records" type="list" label="${tool.name} detected Biosynthetic Gene Clusters on ${on_string} (GenBank)"> 59 <collection name="records" type="list" label="${tool.name} detected Biosynthetic Gene Clusters on ${on_string} (GenBank)">
57 <discover_datasets pattern="(?P&lt;designation&gt;.*)\.gbk" ext="genbank" visible="false" /> 60 <discover_datasets pattern="(?P&lt;designation&gt;.*)\.gbk" ext="genbank" visible="false" />
58 </collection> 61 </collection>
62 <data name="genes" format="tabular" label="${tool.name} summary of detected genes on ${on_string} (TSV)"/>
59 <data name="features" format="tabular" label="${tool.name} summary of detected features on ${on_string} (TSV)"/> 63 <data name="features" format="tabular" label="${tool.name} summary of detected features on ${on_string} (TSV)"/>
60 <data name="clusters" format="tabular" label="${tool.name} summary of detected BGCs on ${on_string} (TSV)"/> 64 <data name="clusters" format="tabular" label="${tool.name} summary of detected BGCs on ${on_string} (TSV)"/>
61 <data name="sideload" format="json" label="antiSMASH v6 sideload file with ${tool.name} detected BGCs on ${on_string} (JSON)"> 65 <data name="sideload" format="json" label="antiSMASH v6 sideload file with ${tool.name} detected BGCs on ${on_string} (JSON)">
62 <filter>antismash_sideload</filter> 66 <filter>antismash_sideload</filter>
63 </data> 67 </data>
64 </outputs> 68 </outputs>
65 <tests> 69 <tests>
66 <test> 70 <test>
67 <param name="input" value="BGC0001866.fna"/> 71 <param name="input" value="BGC0001866.fna"/>
68 <output name="features" file="features.tsv"/> 72 <output name="features" file="features.tsv"/>
73 <output name="genes" file="genes.tsv"/>
69 <output name="clusters" file="clusters.tsv"/> 74 <output name="clusters" file="clusters.tsv"/>
70 </test> 75 </test>
71 <test> 76 <test>
72 <param name="input" value="BGC0001866.fna"/> 77 <param name="input" value="BGC0001866.fna"/>
73 <param name="edge_distance" value="0"/> 78 <param name="edge_distance" value="0"/>
74 <output name="features" file="features.tsv"/> 79 <output name="features" file="features.tsv"/>
80 <output name="genes" file="genes.tsv"/>
75 <output name="clusters" file="clusters.tsv"/> 81 <output name="clusters" file="clusters.tsv"/>
76 <output_collection name="records" type="list"> 82 <output_collection name="records" type="list">
77 <element name="BGC0001866.1_cluster_1" file="BGC0001866.1_cluster_1.gbk" ftype="genbank" compare="diff" lines_diff="4"/> 83 <element name="BGC0001866.1_cluster_1" file="BGC0001866.1_cluster_1.gbk" ftype="genbank" compare="diff" lines_diff="4"/>
78 </output_collection> 84 </output_collection>
79 </test> 85 </test>
80 <test> 86 <test>
81 <param name="input" value="BGC0001866.fna"/> 87 <param name="input" value="BGC0001866.fna"/>
82 <param name="antismash_sideload" value="True"/> 88 <param name="antismash_sideload" value="True"/>
83 <param name="edge_distance" value="0"/> 89 <param name="edge_distance" value="0"/>
84 <output name="features" file="features.tsv"/> 90 <output name="features" file="features.tsv"/>
91 <output name="genes" file="genes.tsv"/>
85 <output name="clusters" file="clusters.tsv"/> 92 <output name="clusters" file="clusters.tsv"/>
86 <output name="sideload" file="sideload.json"/> 93 <output name="sideload" file="sideload.json"/>
87 <output_collection name="records" type="list"> 94 <output_collection name="records" type="list">
88 <element name="BGC0001866.1_cluster_1" file="BGC0001866.1_cluster_1.gbk" ftype="genbank" compare="diff" lines_diff="4"/> 95 <element name="BGC0001866.1_cluster_1" file="BGC0001866.1_cluster_1.gbk" ftype="genbank" compare="diff" lines_diff="4"/>
89 </output_collection> 96 </output_collection>
105 Output 112 Output
106 ------ 113 ------
107 114
108 GECCO will create the following files once done (using the same prefix as the input file): 115 GECCO will create the following files once done (using the same prefix as the input file):
109 116
110 - ``features.tsv``: The features file, containing the identified proteins and domains in the input sequences. 117 - ``features.tsv``: The genes file, containing the genes identified in the input sequences.
111 - ``clusters.tsv``: If any were found, a clusters file, containing the coordinates of the predicted clusters, along their putative biosynthetic type. 118 - ``features.tsv``: The features file, containing the protein domains identified in the input sequences.
119 - ``clusters.tsv``: A clusters file, containing the coordinates of the predicted clusters, along their putative biosynthetic type.
112 - ``{sequence}_cluster_{N}.gbk``: If any BGCs were found, a GenBank file per cluster, containing the cluster sequence annotated with its member proteins and domains. 120 - ``{sequence}_cluster_{N}.gbk``: If any BGCs were found, a GenBank file per cluster, containing the cluster sequence annotated with its member proteins and domains.
113 121
114 Contact 122 Contact
115 ------- 123 -------
116 124