Mercurial > repos > iuc > bcftools_stats

comparison bcftools_stats.xml @ 7:baa8823f8073 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/bcftools commit 2684e1443f03bfe2ae20c31d23817415ec8f7e69
author iuc
date Thu, 21 Feb 2019 15:56:16 -0500
parents 3afcc6474c79
children e3c64225cbe6
comparison
equal deleted inserted replaced
6:b641062a2587 7:baa8823f8073
1 <?xml version='1.0' encoding='utf-8'?> 1 <?xml version='1.0' encoding='utf-8'?>
2 <tool name="bcftools @EXECUTABLE@" id="bcftools_@EXECUTABLE@" version="@VERSION@"> 2 <tool name="bcftools @EXECUTABLE@" id="bcftools_@EXECUTABLE@" version="@TOOL_VERSION@">
3 <description>Parses VCF or BCF and produces stats which can be plotted using plot-vcfstats</description> 3 <description>Parses VCF or BCF and produces stats which can be plotted using plot-vcfstats</description>
4 <macros> 4 <macros>
5 <token name="@EXECUTABLE@">stats</token> 5 <token name="@EXECUTABLE@">stats</token>
6 <import>macros.xml</import> 6 <import>macros.xml</import>
7 </macros> 7 </macros>
8 <expand macro="requirements"> 8 <expand macro="requirements">
9 <expand macro="samtools_requirement"/> 9 <expand macro="samtools_requirement"/>
10 <requirement type="package" version="2.0.0">matplotlib</requirement> 10 <requirement type="package" version="3.0.2">matplotlib</requirement>
11 <requirement type="package" version="0.1.11">tectonic</requirement>
11 </expand> 12 </expand>
12 <expand macro="version_command" /> 13 <expand macro="version_command" />
13 <command detect_errors="aggressive"><![CDATA[ 14 <command detect_errors="aggressive"><![CDATA[
14 @PREPARE_ENV@ 15 @PREPARE_ENV@
16 #set $input_files = [$input_file]
17 #if $inputB_file:
18 #silent $input_files.append($inputB_file)
19 #end if
15 @PREPARE_INPUT_FILES@ 20 @PREPARE_INPUT_FILES@
16 #set $section = $sec_restrict 21 #set $section = $sec_restrict
17 @PREPARE_TARGETS_FILE@ 22 @PREPARE_TARGETS_FILE@
18 @PREPARE_REGIONS_FILE@ 23 @PREPARE_REGIONS_FILE@
19 ## Stats section 24 ## Stats section
27 #set $section = $sec_default 32 #set $section = $sec_default
28 @FASTA_REF@ 33 @FASTA_REF@
29 @EXONS_FILE@ 34 @EXONS_FILE@
30 ${section.first_allele_only} 35 ${section.first_allele_only}
31 #if $section.depth.set_depth == 'yes': 36 #if $section.depth.set_depth == 'yes':
32 --depth "${section.depth.depth_min},${section.depth.depth_max},${section.depth.depth_bin_size}" 37 --depth ${section.depth.depth_min},${section.depth.depth_max},${section.depth.depth_bin_size}
33 #end if 38 #end if
34 #if $section.user_tstv: 39 #if $section.user_tstv:
35 --user-tstv "${section.user_tstv}" 40 --user-tstv '${section.user_tstv}'
36 #end if 41 #end if
37 #if $section.afbins.afbins_select == 'af_bins_list': 42 #if $section.afbins.afbins_select == 'af_bins_list':
38 --af-bins $section.afbins.af_bins_list 43 --af-bins '$section.afbins.af_bins_list'
39 #elif $section.afbins.afbins_select == 'af_bins_file': 44 #elif $section.afbins.afbins_select == 'af_bins_file':
40 --af-bins $section.afbins.af_bins_file 45 --af-bins '$section.afbins.af_bins_file'
41 #end if 46 #end if
42 #if $section.af_tag: 47 #if $section.af_tag:
43 --af-tag "${section.af_tag}" 48 --af-tag '${section.af_tag}'
44 #end if 49 #end if
45 #if len($input_vcfs) == 1: 50 #if len($input_vcfs) == 1:
46 ${section.split_by_ID} 51 ${section.split_by_ID}
47 #end if 52 #end if
48 ${section.verbose} 53 ${section.verbose}
49 54
50 ## Stats section 55 ## Stats section
51 #set $section = $sec_restrict 56 #set $section = $sec_restrict
57 @INCLUDE@ 62 @INCLUDE@
58 @EXCLUDE@ 63 @EXCLUDE@
59 64
60 ## Primary Input/Outputs 65 ## Primary Input/Outputs
61 @INPUT_FILES@ 66 @INPUT_FILES@
62 > $output_file 67 > '$output_file'
63 ## requires: matplotlib and pdflatex
64 #if $plot_title: 68 #if $plot_title:
65 && plot-vcfstats -p 'plot_tmp/' -T $plot_title -s $output_file 69 && plot-vcfstats
70 -p 'plot_tmp/'
71 -T '$plot_title'
72 -s
73 '$output_file'
74 || (printf "The content of plot_tmp/plot-vcfstats.log is:\n" >&2 && cat plot_tmp/plot-vcfstats.log >&2 && exit 1)
66 #end if 75 #end if
67 ]]> 76 ]]></command>
68 </command>
69 <inputs> 77 <inputs>
70 <expand macro="macro_inputs" /> 78 <expand macro="macro_input" />
79 <param name="inputB_file" type="data" format="vcf,vcf_bgzip,bcf" optional="true" label="Optional VCF/BCF Data to compare against" help="When this second dataset is also specified, separate stats for intersection and the complements are generated" />
71 <section name="sec_restrict" expanded="false" title="Restrict to"> 80 <section name="sec_restrict" expanded="false" title="Restrict to">
72 <expand macro="macro_samples" /> 81 <expand macro="macro_samples" />
73 <expand macro="macro_apply_filters" /> 82 <expand macro="macro_apply_filters" />
74 <expand macro="macro_collapse" /> 83 <expand macro="macro_collapse" />
75 <expand macro="macro_regions" /> 84 <expand macro="macro_regions" />
126 <validator type="regex" message="">^\w.*\w$</validator> 135 <validator type="regex" message="">^\w.*\w$</validator>
127 </param> 136 </param>
128 </inputs> 137 </inputs>
129 <outputs> 138 <outputs>
130 <data name="output_file" format="txt"/> 139 <data name="output_file" format="txt"/>
131 <data name="output_pdf" format="pdf"> 140 <data name="output_pdf" format="pdf" from_work_dir="plot_tmp/summary.pdf">
132 <filter>plot_title</filter> 141 <filter>plot_title</filter>
133 </data> 142 </data>
134 </outputs> 143 </outputs>
135 <tests> 144 <tests>
136 <test> 145 <test>
137 <param name="input_files" ftype="vcf" value="stats.b.vcf,stats.a.vcf" /> 146 <param name="input_file" ftype="vcf" value="stats.b.vcf" />
147 <param name="inputB_file" ftype="vcf" value="stats.a.vcf" />
138 <output name="output_file"> 148 <output name="output_file">
139 <assert_contents> 149 <assert_contents>
140 <has_text_matching expression="SN\t0\tnumber of samples:\t3"/> 150 <has_text_matching expression="SN\t0\tnumber of samples:\t3"/>
141 <has_text_matching expression="SN\t1\tnumber of samples:\t3"/> 151 <has_text_matching expression="SN\t1\tnumber of samples:\t3"/>
142 </assert_contents> 152 </assert_contents>
143 </output> 153 </output>
144 </test> 154 </test>
145 <test> 155 <test>
146 <param name="input_files" ftype="vcf" value="mpileup.vcf" /> 156 <param name="input_file" ftype="vcf" value="mpileup.vcf" />
157 <param name="plot_title" value="Plot for mpileup.vcf" />
147 <output name="output_file"> 158 <output name="output_file">
148 <assert_contents> 159 <assert_contents>
149 <has_text_matching expression="SN\t0\tnumber of samples:\t3"/> 160 <has_text_matching expression="SN\t0\tnumber of samples:\t3"/>
150 <has_text_matching expression="SN\t0\tnumber of records:\t4103"/> 161 <has_text_matching expression="SN\t0\tnumber of records:\t4103"/>
151 <has_text_matching expression="ST\t0\tA>C\t16"/> 162 <has_text_matching expression="ST\t0\tA>C\t16"/>
152 </assert_contents> 163 </assert_contents>
153 </output> 164 </output>
165 <output name="output_pdf" file="summary.pdf" compare="sim_size" delta="20000" />
154 </test> 166 </test>
155 </tests> 167 </tests>
156 <help><![CDATA[ 168 <help><![CDATA[
157 ===================================== 169 =====================================
158 bcftools @EXECUTABLE@ 170 bcftools @EXECUTABLE@
160 172
161 Parses VCF or BCF and produces stats which can be plotted using plot-vcfstats. 173 Parses VCF or BCF and produces stats which can be plotted using plot-vcfstats.
162 174
163 When two files are given, the program generates separate stats for intersection and the complements. 175 When two files are given, the program generates separate stats for intersection and the complements.
164 By default only sites are compared, -s/-S must given to include also sample columns. 176 By default only sites are compared, -s/-S must given to include also sample columns.
177
178 When one VCF file is specified, then stats by non-reference allele frequency, depth distribution, stats by quality and per-sample counts, singleton stats, etc. are printed. When two VCF files are given, then stats such as concordance (Genotype concordance by non-reference allele frequency, Genotype concordance by sample, Non-Reference Discordance) and correlation are also printed. Per-site discordance (PSD) is also printed in --verbose mode.
165 179
166 @COLLAPSE_HELP@ 180 @COLLAPSE_HELP@
167 @REGIONS_HELP@ 181 @REGIONS_HELP@
168 @TARGETS_HELP@ 182 @TARGETS_HELP@
169 @EXPRESSIONS_HELP@ 183 @EXPRESSIONS_HELP@