Mercurial > repos > jjohnson > bcftools_annotate

comparison bcftools_annotate.xml @ 0:a30804010531 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 21c66fb27c7e2fd21c7f7607b3b29e77e64fb86d-dirty
author jjohnson
date Sat, 25 Jun 2016 20:46:09 -0400
parents
children 6c6a06ac35c1
comparison
equal deleted inserted replaced
-1:000000000000 0:a30804010531
1 <?xml version='1.0' encoding='utf-8'?>
2 <tool name="bcftools @EXECUTABLE@" id="bcftools_@EXECUTABLE@" version="@VERSION@.0">
3 <description>Annotate and edit VCF/BCF files</description>
4 <macros>
5 <token name="@EXECUTABLE@">annotate</token>
6 <import>macros.xml</import>
7 </macros>
8 <expand macro="requirements" />
9 <expand macro="version_command" />
10 <command detect_errors="aggressive"><![CDATA[
11 @PREPARE_ENV@
12 @PREPARE_INPUT_FILE@
13 #set $annotation_file = None
14 #set $annotation_hdr = None
15 #set $section = $sec_annofile
16 #if $section.annofile.anno_fmt == 'vcf':
17 #set $annotation_file = 'annotations.vcf.gz'
18 bgzip -c "$section.annofile.annotations" > $annotation_file &&
19 bcftools index $annotation_file &&
20 #elif $section.annofile.anno_fmt == 'tab':
21 #if $section.annofile.annotations.ext == 'bed'
22 #set $annotation_file = 'annotations.bed.gz'
23 bgzip -c "$section.annofile.annotations" > $annotation_file &&
24 tabix -s 1 -b 2 -e 3 $annotation_file &&
25 #else:
26 #set $annotation_file = 'annotations.tab.gz'
27 bgzip -c "$section.annofile.annotations" > $annotation_file &&
28 tabix -s 1 -b 2 -e 2 $annotation_file &&
29 #end if
30 #if $section.annofile.header_file:
31 #set $annotation_hdr = $section.annofile.header_file
32 #elif $section.annofile.header_lines:
33 #set $annotation_hdr = 'annotation.hdr'
34 grep '^\#\#INFO' ${hdr_file} > $annotation_hdr &&
35 #end if
36 #end if
37
38 bcftools @EXECUTABLE@
39
40 #set $section = $sec_annofile
41 @COLUMNS@
42 #if $annotation_file:
43 --annotations "${annotation_file}"
44 #end if
45 #if $annotation_hdr:
46 --header-lines "${annotation_hdr}"
47 #end if
48
49 #if $section.set_id:
50 --set-id "${section.set_id}"
51 #end if
52 #if $section.mark_sites:
53 --mark-sites "${section.mark_sites}"
54 #end if
55
56 #set $section = $sec_annotate
57 #if $section.rename_chrs:
58 --rename-chrs "${section.rename_chrs}"
59 #end if
60 #if $section.remove:
61 --remove "${section.remove}"
62 #end if
63
64 ## Default section
65 #set $section = $sec_restrict
66 @INCLUDE@
67 @EXCLUDE@
68 @REGIONS@
69 @SAMPLES@
70
71 @OUTPUT_TYPE@
72 @THREADS@
73
74 ## Primary Input/Outputs
75 @INPUT_FILE@
76 > "$output_file"
77 ]]>
78 </command>
79 <configfiles>
80 <configfile name="hdr_file"><![CDATA[#slurp
81 #if $sec_annofile.annofile.anno_fmt == 'tab' and str($sec_annofile.annofile.header_lines) != '':
82 $sec_annofile.annofile.header_lines.__str__.strip()#slurp
83 #end if]]></configfile>
84 </configfiles>
85 <inputs>
86 <expand macro="macro_input" />
87 <section name="sec_annofile" expanded="false" title="Add Annotations">
88 <expand macro="macro_columns" />
89 <conditional name="annofile">
90 <param name="anno_fmt" type="select" label="Annotations File">
91 <option value="none">None</option>
92 <option value="vcf">From a VCF/BCF file</option>
93 <option value="tab">From a BED or tab-delimited file</option>
94 </param>
95 <when value="none"/>
96 <when value="vcf">
97 <param name="annotations" type="data" format="vcf" label="Annotations VCF"/>
98 </when>
99 <when value="tab">
100 <param name="annotations" type="data" format="tabular,bed" label="Annotations">
101 <help><![CDATA[
102 BED, or a tab-delimited file with mandatory columns CHROM, POS (or, alternatively, FROM and TO),
103 optional columns REF and ALT, and arbitrary number of annotation columns.
104 Note that in case of tab-delimited file, the coordinates POS, FROM and TO are one-based and inclusive.
105 ]]></help>
106 </param>
107 <param name="header_file" type="data" format="txt" label="Header Lines File" optional="True" help="lines which should be appended to the VCF header" />
108 <param name="header_lines" type="text" area="True" label="Header Lines" optional="True" help="lines which should be appended to the VCF header" >
109 <help><![CDATA[
110 ##INFO=<ID=NUMERIC_TAG,Number=1,Type=Integer,Description="Example header line">
111 ##INFO=<ID=STRING_TAG,Number=1,Type=String,Description="Yet another header line">
112 ]]></help>
113 <sanitizer sanitize="False"/>
114 </param>
115 </when>
116 </conditional>
117 <param name="mark_sites" type="text" value="" label="Mark Sites TAG"
118 help="add INFO/TAG flag to sites which are (&quot;+&quot;) or are not (&quot;-&quot;) listed in the annotations file" />
119 <param name="set_id" type="text" value="" optional="true" label="Set Id">
120 <help>Assign ID on the fly using the given format.
121 By default all existing IDs are replaced.
122 If the format string is preceded by "+", only missing IDs will be set.
123 For example: '%CHROM\_%POS\_%REF\_%FIRST_ALT'
124 </help>
125 <sanitizer sanitize="False"/>
126 <validator type="regex" message="">^([+]?(%[A-Z]+)(\_%[A-Z]+)*)?$</validator>
127 </param>
128 </section>
129 <section name="sec_annotate" expanded="false" title="Change Annotations">
130 <param name="remove" type="text" value="" label="Remove annotations" optional="true">
131 <help><![CDATA[
132 List of annotations to remove.
133 Use "FILTER" to remove all filters or "FILTER/SomeFilter" to remove a specific filter.
134 Similarly, "INFO" can be used to remove all INFO tags and "FORMAT" to remove all FORMAT tags except GT.
135 To remove all INFO tags except "FOO" and "BAR", use "^INFO/FOO,INFO/BAR" (and similarly for FORMAT and FILTER).
136 "INFO" can be abbreviated to "INF" and "FORMAT" to "FMT".
137 ]]></help>
138 <validator type="regex" message="">^([+]?(%[A-Z]+)(\_%[A-Z]+)*)?$</validator>
139 </param>
140 <param name="rename_chrs" type="data" format="tabular" label="Rename CHROM" optional="True"
141 help="rename chromosomes according to the map in file, with old_name new_name pairs separated by whitespaces, each on a separate line." />
142 </section>
143 <section name="sec_restrict" expanded="false" title="Restrict to">
144 <expand macro="macro_include" />
145 <expand macro="macro_exclude" />
146 <expand macro="macro_regions" />
147 <expand macro="macro_samples" />
148 </section>
149 <expand macro="macro_select_output_type" />
150 </inputs>
151 <outputs>
152 <expand macro="macro_vcf_output"/>
153 </outputs>
154 <tests>
155 <test>
156 <param name="input_file" ftype="vcf" value="annotate.vcf" />
157 <param name="anno_fmt" value="tab" />
158 <param name="annotations" value="annotate.tab" />
159 <param name="header_file" value="annotate.hdr" />
160 <param name="columns" value="CHROM,POS,REF,ALT,ID,QUAL,INFO/T_INT,INFO/T_FLOAT,INDEL" />
161 <param name="output_type" value="v" />
162 <output name="output_file">
163 <assert_contents>
164 <has_text text="snp_3000150" />
165 </assert_contents>
166 </output>
167 </test>
168 <test>
169 <param name="input_file" ftype="vcf" value="annotate.vcf" />
170 <param name="anno_fmt" value="tab" />
171 <param name="annotations" value="annotate2.tab" />
172 <param name="header_file" value="annotate.hdr" />
173 <param name="columns" value="CHROM,FROM,TO,T_STR" />
174 <param name="output_type" value="v" />
175 <output name="output_file">
176 <assert_contents>
177 <has_text text="T_STR=region_3000150_3106154" />
178 </assert_contents>
179 </output>
180 </test>
181 <test>
182 <param name="input_file" ftype="vcf" value="annotate.vcf" />
183 <param name="anno_fmt" value="vcf" />
184 <param name="annotations" value="annots.vcf" />
185 <param name="columns" value="STR,ID,QUAL,FILTER" />
186 <param name="output_type" value="v" />
187 <output name="output_file">
188 <assert_contents>
189 <has_text text="idIndel" />
190 <has_text text="STR=testSNP" />
191 </assert_contents>
192 </output>
193 </test>
194 <test>
195 <param name="input_file" ftype="vcf" value="annotate2.vcf" />
196 <param name="anno_fmt" value="vcf" />
197 <param name="annotations" value="annots2.vcf" />
198 <param name="columns" value="ID,QUAL,FILTER,INFO,FMT" />
199 <param name="output_type" value="v" />
200 <output name="output_file">
201 <assert_contents>
202 <has_text text="q99" />
203 <has_text text="FLAG;IINT=88,99;IFLT=8.8,9.9;ISTR=888,999" />
204 </assert_contents>
205 </output>
206 </test>
207 <test>
208 <param name="input_file" ftype="vcf" value="annotate2.vcf" />
209 <param name="anno_fmt" value="vcf" />
210 <param name="annotations" value="annots2.vcf" />
211 <param name="columns" value="ID,QUAL,FILTER,INFO,FMT" />
212 <param name="samples" value="A" />
213 <param name="output_type" value="v" />
214 <output name="output_file">
215 <assert_contents>
216 <has_text text="q11" />
217 <has_text text="FLAG;IINT=88,99;IFLT=8.8,9.9;ISTR=888,999" />
218 </assert_contents>
219 </output>
220 </test>
221 <test>
222 <param name="input_file" ftype="vcf" value="annotate3.vcf" />
223 <param name="anno_fmt" value="none" />
224 <param name="remove" value="ID,QUAL,^FILTER/fltA,FILTER/fltB,^INFO/AA,INFO/BB,^FMT/GT,FMT/PL" />
225 <param name="output_type" value="v" />
226 <output name="output_file">
227 <assert_contents>
228 <not_has_text text="ID=X" />
229 </assert_contents>
230 </output>
231 </test>
232 <test>
233 <param name="input_file" ftype="vcf" value="annotate3.vcf" />
234 <param name="anno_fmt" value="none" />
235 <param name="remove" value="FORMAT" />
236 <param name="output_type" value="v" />
237 <output name="output_file">
238 <assert_contents>
239 <not_has_text text="ID=PL" />
240 </assert_contents>
241 </output>
242 </test>
243
244 </tests>
245 <help><![CDATA[
246 ==================================
247 bcftools @EXECUTABLE@
248 ==================================
249
250 Annotate and edit VCF/BCF files.
251
252 Examples:
253
254 # Remove three fields
255 bcftools annotate -x ID,INFO/DP,FORMAT/DP file.vcf.gz
256
257 # Remove all INFO fields and all FORMAT fields except for GT and PL
258 bcftools annotate -x INFO,^FORMAT/GT,FORMAT/PL file.vcf
259
260 # Add ID, QUAL and INFO/TAG, not replacing TAG if already present
261 bcftools annotate -a src.bcf -c ID,QUAL,+TAG dst.bcf
262
263 # Carry over all INFO and FORMAT annotations except FORMAT/GT
264 bcftools annotate -a src.bcf -c INFO,^FORMAT/GT dst.bcf
265
266 # Annotate from a tab-delimited file with six columns (the fifth is ignored),
267 # first indexing with tabix. The coordinates are 1-based.
268 tabix -s1 -b2 -e2 annots.tab.gz
269 bcftools annotate -a annots.tab.gz -h annots.hdr -c CHROM,POS,REF,ALT,-,TAG file.vcf
270
271 # Annotate from a tab-delimited file with regions (1-based coordinates, inclusive)
272 tabix -s1 -b2 -e3 annots.tab.gz
273 bcftools annotate -a annots.tab.gz -h annots.hdr -c CHROM,FROM,TO,TAG inut.vcf
274
275 # Annotate from a bed file (0-based coordinates, half-closed, half-open intervals)
276 bcftools annotate -a annots.bed.gz -h annots.hdr -c CHROM,FROM,TO,TAG input.vcf
277
278 @REGIONS_HELP@
279 @EXPRESSIONS_HELP@
280
281 @BCFTOOLS_MANPAGE@#@EXECUTABLE@
282
283 @BCFTOOLS_WIKI@
284 ]]>
285 </help>
286 <expand macro="citations" />
287 </tool>