|
0
|
1 <tool id="gatk2_variant_combine" name="Combine Variants" version="0.0.4">
|
|
|
2 <description></description>
|
|
|
3 <requirements>
|
|
|
4 <requirement type="package" version="2.2">gatk</requirement>
|
|
|
5 </requirements>
|
|
|
6 <command interpreter="python">gatk2_wrapper.py
|
|
|
7 --max_jvm_heap_fraction "1"
|
|
|
8 --stdout "${output_log}"
|
|
|
9
|
|
|
10 #set $priority_order = []
|
|
|
11 #for $input_variant in $reference_source.input_variants:
|
|
|
12 -d "--variant:${input_variant.input_variant_name},%(file_type)s" "${input_variant.input_variant}" "${input_variant.input_variant.ext}" "input_variant_${input_variant.input_variant_name}"
|
|
|
13 #set $input_variant_name = str( $input_variant.input_variant_name )
|
|
|
14 #assert $input_variant_name not in $priority_order, "Variant Names must be unique" ##this should be handled by a validator
|
|
|
15 #silent $priority_order.append( $input_variant_name )
|
|
|
16 #end for
|
|
|
17 -p 'java
|
|
|
18 -jar "\$GATK2_PATH/GenomeAnalysisTK.jar"
|
|
|
19 -T "CombineVariants"
|
|
|
20 --out "${output_variants}"
|
|
|
21 \$GATK2_SITE_OPTIONS
|
|
|
22 ##--num_threads 4 ##hard coded, for now
|
|
|
23 ##-et "NO_ET" -K "\$GATK2_BASE/gatk2_key_file" ##ET no phone home
|
|
|
24 ##-log "${output_log}" ##don't use this to log to file, instead directly capture stdout
|
|
|
25 #if $reference_source.reference_source_selector != "history":
|
|
|
26 -R "${reference_source.ref_file.fields.path}"
|
|
|
27 #end if
|
|
|
28 --genotypemergeoption "${genotype_merge_option}"
|
|
|
29 --rod_priority_list "${ ','.join( $priority_order ) }"
|
|
|
30 '
|
|
|
31
|
|
|
32 ##start standard gatk options
|
|
|
33 #if $gatk_param_type.gatk_param_type_selector == "advanced":
|
|
|
34 #for $pedigree in $gatk_param_type.pedigree:
|
|
|
35 -p '--pedigree "${pedigree.pedigree_file}"'
|
|
|
36 #end for
|
|
|
37 #for $pedigree_string in $gatk_param_type.pedigree_string_repeat:
|
|
|
38 -p '--pedigreeString "${pedigree_string.pedigree_string}"'
|
|
|
39 #end for
|
|
|
40 -p '--pedigreeValidationType "${gatk_param_type.pedigree_validation_type}"'
|
|
|
41 #for $read_filter in $gatk_param_type.read_filter:
|
|
|
42 -p '--read_filter "${read_filter.read_filter_type.read_filter_type_selector}"
|
|
|
43 ###raise Exception( str( dir( $read_filter ) ) )
|
|
|
44 #for $name, $param in $read_filter.read_filter_type.iteritems():
|
|
|
45 #if $name not in [ "__current_case__", "read_filter_type_selector" ]:
|
|
|
46 #if hasattr( $param.input, 'truevalue' ):
|
|
|
47 ${param}
|
|
|
48 #else:
|
|
|
49 --${name} "${param}"
|
|
|
50 #end if
|
|
|
51 #end if
|
|
|
52 #end for
|
|
|
53 '
|
|
|
54 #end for
|
|
|
55 #for $interval_count, $input_intervals in enumerate( $gatk_param_type.input_interval_repeat ):
|
|
|
56 -d "--intervals" "${input_intervals.input_intervals}" "${input_intervals.input_intervals.ext}" "input_intervals_${interval_count}"
|
|
|
57 #end for
|
|
|
58
|
|
|
59 #for $interval_count, $input_intervals in enumerate( $gatk_param_type.input_exclude_interval_repeat ):
|
|
|
60 -d "--excludeIntervals" "${input_intervals.input_exclude_intervals}" "${input_intervals.input_exclude_intervals.ext}" "input_exlude_intervals_${interval_count}"
|
|
|
61 #end for
|
|
|
62
|
|
|
63 -p '--interval_set_rule "${gatk_param_type.interval_set_rule}"'
|
|
|
64
|
|
|
65 -p '--downsampling_type "${gatk_param_type.downsampling_type.downsampling_type_selector}"'
|
|
|
66 #if str( $gatk_param_type.downsampling_type.downsampling_type_selector ) != "NONE":
|
|
|
67 -p '--${gatk_param_type.downsampling_type.downsample_to_type.downsample_to_type_selector} "${gatk_param_type.downsampling_type.downsample_to_type.downsample_to_value}"'
|
|
|
68 #end if
|
|
|
69 -p '
|
|
|
70 --baq "${gatk_param_type.baq}"
|
|
|
71 --baqGapOpenPenalty "${gatk_param_type.baq_gap_open_penalty}"
|
|
|
72 ${gatk_param_type.use_original_qualities}
|
|
|
73 --defaultBaseQualities "${gatk_param_type.default_base_qualities}"
|
|
|
74 --validation_strictness "${gatk_param_type.validation_strictness}"
|
|
|
75 --interval_merging "${gatk_param_type.interval_merging}"
|
|
|
76 ${gatk_param_type.disable_experimental_low_memory_sharding}
|
|
|
77 ${gatk_param_type.non_deterministic_random_seed}
|
|
|
78 '
|
|
|
79 #for $rg_black_list_count, $rg_black_list in enumerate( $gatk_param_type.read_group_black_list_repeat ):
|
|
|
80 #if $rg_black_list.read_group_black_list_type.read_group_black_list_type_selector == "file":
|
|
|
81 -d "--read_group_black_list" "${rg_black_list.read_group_black_list_type.read_group_black_list}" "txt" "input_read_group_black_list_${rg_black_list_count}"
|
|
|
82 #else
|
|
|
83 -p '--read_group_black_list "${rg_black_list.read_group_black_list_type.read_group_black_list}"'
|
|
|
84 #end if
|
|
|
85 #end for
|
|
|
86 #end if
|
|
|
87
|
|
|
88 #if $reference_source.reference_source_selector == "history":
|
|
|
89 -d "-R" "${reference_source.ref_file}" "${reference_source.ref_file.ext}" "gatk_input"
|
|
|
90 #end if
|
|
|
91 ##end standard gatk options
|
|
|
92
|
|
|
93
|
|
|
94 ##start analysis specific options
|
|
|
95 #if $analysis_param_type.analysis_param_type_selector == "advanced":
|
|
|
96 -p '
|
|
|
97 --filteredrecordsmergetype "${analysis_param_type.filtered_records_merge_type}"
|
|
|
98 ${analysis_param_type.print_complex_merges}
|
|
|
99 ${analysis_param_type.filtered_are_uncalled}
|
|
|
100 ${analysis_param_type.minimal_vcf}
|
|
|
101 ${analysis_param_type.assume_identical_samples}
|
|
|
102
|
|
|
103 #if str( $analysis_param_type.set_key ):
|
|
|
104 --setKey "${analysis_param_type.set_key}"
|
|
|
105 #end if
|
|
|
106
|
|
|
107 --minimumN "${analysis_param_type.minimum_n}"
|
|
|
108 '
|
|
|
109 #end if
|
|
|
110 </command>
|
|
|
111 <inputs>
|
|
|
112
|
|
|
113 <conditional name="reference_source">
|
|
|
114 <param name="reference_source_selector" type="select" label="Choose the source for the reference list">
|
|
|
115 <option value="cached">Locally cached</option>
|
|
|
116 <option value="history">History</option>
|
|
|
117 </param>
|
|
|
118 <when value="cached">
|
|
|
119 <repeat min="1" name="input_variants" title="Variants to Merge" help="Records will be prioritized in the order that you list them here (-V,--variant &lt;variant&gt;)">
|
|
|
120 <param name="input_variant" type="data" format="vcf" label="Input variant file" />
|
|
|
121 <param name="input_variant_name" type="text" value="" label="Variant name" help="Names must be unique">
|
|
|
122 <validator type="length" min="1" message="You must provide a unique name for this set of variants" />
|
|
|
123 </param>
|
|
|
124 </repeat>
|
|
|
125 <param name="ref_file" type="select" label="Using reference genome" help="-R,--reference_sequence &lt;reference_sequence&gt;">
|
|
|
126 <options from_data_table="gatk2_picard_indexes">
|
|
|
127 <!-- <filter type="data_meta" key="dbkey" ref="input_variants.input_variant" column="dbkey"/> -->
|
|
|
128 </options>
|
|
|
129 <validator type="no_options" message="A built-in reference genome is not available for the build associated with the selected input file"/>
|
|
|
130 </param>
|
|
|
131 </when>
|
|
|
132 <when value="history"> <!-- FIX ME!!!! -->
|
|
|
133 <repeat min="1" name="input_variants" title="Variants to Merge" help="Records will be prioritized in the order that you list them here (-V,--variant &lt;variant&gt;)">
|
|
|
134 <param name="input_variant" type="data" format="vcf" label="Input variant file" />
|
|
|
135 <param name="input_variant_name" type="text" value="" label="Variant name" help="Names must be unique">
|
|
|
136 <validator type="length" min="1" message="You must provide a unique name for this set of variants" />
|
|
|
137 </param>
|
|
|
138 </repeat>
|
|
|
139 <param name="ref_file" type="data" format="fasta" label="Using reference file" help="-R,--reference_sequence &lt;reference_sequence&gt;" />
|
|
|
140 </when>
|
|
|
141 </conditional>
|
|
|
142
|
|
|
143 <param name="genotype_merge_option" type="select" label="How should we merge genotype records across records for samples shared across the ROD files" help="-genotypeMergeOptions,--genotypemergeoption &lt;genotypemergeoption&gt;" >
|
|
|
144 <option value="UNIQUIFY" />
|
|
|
145 <option value="PRIORITIZE" selected="true"/>
|
|
|
146 <option value="UNSORTED" />
|
|
|
147 <option value="REQUIRE_UNIQUE" />
|
|
|
148 </param>
|
|
|
149
|
|
|
150 <conditional name="gatk_param_type">
|
|
|
151 <param name="gatk_param_type_selector" type="select" label="Basic or Advanced GATK options">
|
|
|
152 <option value="basic" selected="True">Basic</option>
|
|
|
153 <option value="advanced">Advanced</option>
|
|
|
154 </param>
|
|
|
155 <when value="basic">
|
|
|
156 <!-- Do nothing here -->
|
|
|
157 </when>
|
|
|
158 <when value="advanced">
|
|
|
159 <repeat name="pedigree" title="Pedigree file" help="-ped,--pedigree &lt;pedigree&gt;">
|
|
|
160 <param name="pedigree_file" type="data" format="txt" label="Pedigree files for samples"/>
|
|
|
161 </repeat>
|
|
|
162 <repeat name="pedigree_string_repeat" title="Pedigree string" help="-pedString,--pedigreeString &lt;pedigreeString&gt;">
|
|
|
163 <param name="pedigree_string" type="text" value="" label="Pedigree string for samples"/>
|
|
|
164 </repeat>
|
|
|
165 <param name="pedigree_validation_type" type="select" label="How strict should we be in validating the pedigree information" help="-pedValidationType,--pedigreeValidationType &lt;pedigreeValidationType&gt;">
|
|
|
166 <option value="STRICT" selected="True">STRICT</option>
|
|
|
167 <option value="SILENT">SILENT</option>
|
|
|
168 </param>
|
|
|
169 <repeat name="read_filter" title="Read Filter" help="-rf,--read_filter &lt;read_filter&gt;">
|
|
|
170 <conditional name="read_filter_type">
|
|
|
171 <param name="read_filter_type_selector" type="select" label="Read Filter Type">
|
|
|
172 <option value="BadCigar">BadCigar</option>
|
|
|
173 <option value="BadMate">BadMate</option>
|
|
|
174 <option value="DuplicateRead">DuplicateRead</option>
|
|
|
175 <option value="FailsVendorQualityCheck">FailsVendorQualityCheck</option>
|
|
|
176 <option value="MalformedRead">MalformedRead</option>
|
|
|
177 <option value="MappingQuality">MappingQuality</option>
|
|
|
178 <option value="MappingQualityUnavailable">MappingQualityUnavailable</option>
|
|
|
179 <option value="MappingQualityZero">MappingQualityZero</option>
|
|
|
180 <option value="MateSameStrand">MateSameStrand</option>
|
|
|
181 <option value="MaxInsertSize">MaxInsertSize</option>
|
|
|
182 <option value="MaxReadLength" selected="True">MaxReadLength</option>
|
|
|
183 <option value="MissingReadGroup">MissingReadGroup</option>
|
|
|
184 <option value="NoOriginalQualityScores">NoOriginalQualityScores</option>
|
|
|
185 <option value="NotPrimaryAlignment">NotPrimaryAlignment</option>
|
|
|
186 <option value="Platform454">Platform454</option>
|
|
|
187 <option value="Platform">Platform</option>
|
|
|
188 <option value="PlatformUnit">PlatformUnit</option>
|
|
|
189 <option value="ReadGroupBlackList">ReadGroupBlackList</option>
|
|
|
190 <option value="ReadName">ReadName</option>
|
|
|
191 <option value="ReadStrand">ReadStrand</option>
|
|
|
192 <option value="ReassignMappingQuality">ReassignMappingQuality</option>
|
|
|
193 <option value="Sample">Sample</option>
|
|
|
194 <option value="SingleReadGroup">SingleReadGroup</option>
|
|
|
195 <option value="UnmappedRead">UnmappedRead</option>
|
|
|
196 </param>
|
|
|
197 <when value="BadCigar">
|
|
|
198 <!-- no extra options -->
|
|
|
199 </when>
|
|
|
200 <when value="BadMate">
|
|
|
201 <!-- no extra options -->
|
|
|
202 </when>
|
|
|
203 <when value="DuplicateRead">
|
|
|
204 <!-- no extra options -->
|
|
|
205 </when>
|
|
|
206 <when value="FailsVendorQualityCheck">
|
|
|
207 <!-- no extra options -->
|
|
|
208 </when>
|
|
|
209 <when value="MalformedRead">
|
|
|
210 <!-- no extra options -->
|
|
|
211 </when>
|
|
|
212 <when value="MappingQuality">
|
|
|
213 <param name="min_mapping_quality_score" type="integer" value="10" label="Minimum read mapping quality required to consider a read for calling"/>
|
|
|
214 </when>
|
|
|
215 <when value="MappingQualityUnavailable">
|
|
|
216 <!-- no extra options -->
|
|
|
217 </when>
|
|
|
218 <when value="MappingQualityZero">
|
|
|
219 <!-- no extra options -->
|
|
|
220 </when>
|
|
|
221 <when value="MateSameStrand">
|
|
|
222 <!-- no extra options -->
|
|
|
223 </when>
|
|
|
224 <when value="MaxInsertSize">
|
|
|
225 <param name="maxInsertSize" type="integer" value="1000000" label="Discard reads with insert size greater than the specified value"/>
|
|
|
226 </when>
|
|
|
227 <when value="MaxReadLength">
|
|
|
228 <param name="maxReadLength" type="integer" value="76" label="Max Read Length"/>
|
|
|
229 </when>
|
|
|
230 <when value="MissingReadGroup">
|
|
|
231 <!-- no extra options -->
|
|
|
232 </when>
|
|
|
233 <when value="NoOriginalQualityScores">
|
|
|
234 <!-- no extra options -->
|
|
|
235 </when>
|
|
|
236 <when value="NotPrimaryAlignment">
|
|
|
237 <!-- no extra options -->
|
|
|
238 </when>
|
|
|
239 <when value="Platform454">
|
|
|
240 <!-- no extra options -->
|
|
|
241 </when>
|
|
|
242 <when value="Platform">
|
|
|
243 <param name="PLFilterName" type="text" value="" label="Discard reads with RG:PL attribute containing this string"/>
|
|
|
244 </when>
|
|
|
245 <when value="PlatformUnit">
|
|
|
246 <!-- no extra options -->
|
|
|
247 </when>
|
|
|
248 <when value="ReadGroupBlackList">
|
|
|
249 <!-- no extra options -->
|
|
|
250 </when>
|
|
|
251 <when value="ReadName">
|
|
|
252 <param name="readName" type="text" value="" label="Filter out all reads except those with this read name"/>
|
|
|
253 </when>
|
|
|
254 <when value="ReadStrand">
|
|
|
255 <param name="filterPositive" type="boolean" truevalue="--filterPositive" falsevalue="" label="Discard reads on the forward strand"/>
|
|
|
256 </when>
|
|
|
257 <when value="ReassignMappingQuality">
|
|
|
258 <param name="default_mapping_quality" type="integer" value="60" label="Default read mapping quality to assign to all reads"/>
|
|
|
259 </when>
|
|
|
260 <when value="Sample">
|
|
|
261 <param name="sample_to_keep" type="text" value="" label="The name of the sample(s) to keep, filtering out all others"/>
|
|
|
262 </when>
|
|
|
263 <when value="SingleReadGroup">
|
|
|
264 <param name="read_group_to_keep" type="integer" value="76" label="The name of the read group to keep, filtering out all others"/>
|
|
|
265 </when>
|
|
|
266 <when value="UnmappedRead">
|
|
|
267 <!-- no extra options -->
|
|
|
268 </when>
|
|
|
269 </conditional>
|
|
|
270 </repeat>
|
|
|
271 <repeat name="input_interval_repeat" title="Operate on Genomic intervals" help="-L,--intervals &lt;intervals&gt;">
|
|
|
272 <param name="input_intervals" type="data" format="bed,gatk_interval,picard_interval_list,vcf" label="Genomic intervals" />
|
|
|
273 </repeat>
|
|
|
274 <repeat name="input_exclude_interval_repeat" title="Exclude Genomic intervals" help="-XL,--excludeIntervals &lt;excludeIntervals&gt;">
|
|
|
275 <param name="input_exclude_intervals" type="data" format="bed,gatk_interval,picard_interval_list,vcf" label="Genomic intervals" />
|
|
|
276 </repeat>
|
|
|
277
|
|
|
278 <param name="interval_set_rule" type="select" label="Interval set rule" help="-isr,--interval_set_rule &lt;interval_set_rule&gt;">
|
|
|
279 <option value="UNION" selected="True">UNION</option>
|
|
|
280 <option value="INTERSECTION">INTERSECTION</option>
|
|
|
281 </param>
|
|
|
282
|
|
|
283 <conditional name="downsampling_type">
|
|
|
284 <param name="downsampling_type_selector" type="select" label="Type of reads downsampling to employ at a given locus" help="-dt,--downsampling_type &lt;downsampling_type&gt;">
|
|
|
285 <option value="NONE" selected="True">NONE</option>
|
|
|
286 <option value="ALL_READS">ALL_READS</option>
|
|
|
287 <option value="BY_SAMPLE">BY_SAMPLE</option>
|
|
|
288 </param>
|
|
|
289 <when value="NONE">
|
|
|
290 <!-- no more options here -->
|
|
|
291 </when>
|
|
|
292 <when value="ALL_READS">
|
|
|
293 <conditional name="downsample_to_type">
|
|
|
294 <param name="downsample_to_type_selector" type="select" label="Downsample method">
|
|
|
295 <option value="downsample_to_fraction" selected="True">Downsample by Fraction</option>
|
|
|
296 <option value="downsample_to_coverage">Downsample by Coverage</option>
|
|
|
297 </param>
|
|
|
298 <when value="downsample_to_fraction">
|
|
|
299 <param name="downsample_to_value" type="float" label="Fraction [0.0-1.0] of reads to downsample to" value="1" min="0" max="1" help="-dfrac,--downsample_to_fraction &lt;downsample_to_fraction&gt;"/>
|
|
|
300 </when>
|
|
|
301 <when value="downsample_to_coverage">
|
|
|
302 <param name="downsample_to_value" type="integer" label="Coverage to downsample to at any given locus" value="0" help="-dcov,--downsample_to_coverage &lt;downsample_to_coverage&gt;"/>
|
|
|
303 </when>
|
|
|
304 </conditional>
|
|
|
305 </when>
|
|
|
306 <when value="BY_SAMPLE">
|
|
|
307 <conditional name="downsample_to_type">
|
|
|
308 <param name="downsample_to_type_selector" type="select" label="Downsample method">
|
|
|
309 <option value="downsample_to_fraction" selected="True">Downsample by Fraction</option>
|
|
|
310 <option value="downsample_to_coverage">Downsample by Coverage</option>
|
|
|
311 </param>
|
|
|
312 <when value="downsample_to_fraction">
|
|
|
313 <param name="downsample_to_value" type="float" label="Fraction [0.0-1.0] of reads to downsample to" value="1" min="0" max="1" help="-dfrac,--downsample_to_fraction &lt;downsample_to_fraction&gt;"/>
|
|
|
314 </when>
|
|
|
315 <when value="downsample_to_coverage">
|
|
|
316 <param name="downsample_to_value" type="integer" label="Coverage to downsample to at any given locus" value="0" help="-dcov,--downsample_to_coverage &lt;downsample_to_coverage&gt;"/>
|
|
|
317 </when>
|
|
|
318 </conditional>
|
|
|
319 </when>
|
|
|
320 </conditional>
|
|
|
321 <param name="baq" type="select" label="Type of BAQ calculation to apply in the engine" help="-baq,--baq &lt;baq&gt;">
|
|
|
322 <option value="OFF" selected="True">OFF</option>
|
|
|
323 <option value="CALCULATE_AS_NECESSARY">CALCULATE_AS_NECESSARY</option>
|
|
|
324 <option value="RECALCULATE">RECALCULATE</option>
|
|
|
325 </param>
|
|
|
326 <param name="baq_gap_open_penalty" type="float" label="BAQ gap open penalty (Phred Scaled)" value="40" help="Default value is 40. 30 is perhaps better for whole genome call sets. -baqGOP,--baqGapOpenPenalty &lt;baqGapOpenPenalty&gt;" />
|
|
|
327 <param name="use_original_qualities" type="boolean" truevalue="--useOriginalQualities" falsevalue="" label="Use the original base quality scores from the OQ tag" help="-OQ,--useOriginalQualities" />
|
|
|
328 <param name="default_base_qualities" type="integer" label="Value to be used for all base quality scores, when some are missing" value="-1" help="-DBQ,--defaultBaseQualities &lt;defaultBaseQualities&gt;"/>
|
|
|
329 <param name="validation_strictness" type="select" label="How strict should we be with validation" help="-S,--validation_strictness &lt;validation_strictness&gt;">
|
|
|
330 <option value="STRICT" selected="True">STRICT</option>
|
|
|
331 <option value="LENIENT">LENIENT</option>
|
|
|
332 <option value="SILENT">SILENT</option>
|
|
|
333 <!-- <option value="DEFAULT_STRINGENCY">DEFAULT_STRINGENCY</option> listed in docs, but not valid value...-->
|
|
|
334 </param>
|
|
|
335 <param name="interval_merging" type="select" label="Interval merging rule" help="-im,--interval_merging &lt;interval_merging&gt;">
|
|
|
336 <option value="ALL" selected="True">ALL</option>
|
|
|
337 <option value="OVERLAPPING_ONLY">OVERLAPPING_ONLY</option>
|
|
|
338 </param>
|
|
|
339
|
|
|
340 <repeat name="read_group_black_list_repeat" title="Read group black list" help="-rgbl,--read_group_black_list &lt;read_group_black_list&gt;">
|
|
|
341 <conditional name="read_group_black_list_type">
|
|
|
342 <param name="read_group_black_list_type_selector" type="select" label="Type of reads read group black list">
|
|
|
343 <option value="file" selected="True">Filters in file</option>
|
|
|
344 <option value="text">Specify filters as a string</option>
|
|
|
345 </param>
|
|
|
346 <when value="file">
|
|
|
347 <param name="read_group_black_list" type="data" format="txt" label="Read group black list file" />
|
|
|
348 </when>
|
|
|
349 <when value="text">
|
|
|
350 <param name="read_group_black_list" type="text" value="tag:string" label="Read group black list tag:string" />
|
|
|
351 </when>
|
|
|
352 </conditional>
|
|
|
353 </repeat>
|
|
|
354
|
|
|
355 <param name="disable_experimental_low_memory_sharding" type="boolean" truevalue="--disable_experimental_low_memory_sharding" falsevalue="" label="Disable experimental low-memory sharding functionality." checked="False" help="--disable_experimental_low_memory_sharding"/>
|
|
|
356 <param name="non_deterministic_random_seed" type="boolean" truevalue="--nonDeterministicRandomSeed" falsevalue="" label="Makes the GATK behave non deterministically, that is, the random numbers generated will be different in every run" checked="False" help="-ndrs,--nonDeterministicRandomSeed"/>
|
|
|
357
|
|
|
358 </when>
|
|
|
359 </conditional>
|
|
|
360
|
|
|
361
|
|
|
362 <conditional name="analysis_param_type">
|
|
|
363 <param name="analysis_param_type_selector" type="select" label="Basic or Advanced Analysis options">
|
|
|
364 <option value="basic" selected="True">Basic</option>
|
|
|
365 <option value="advanced">Advanced</option>
|
|
|
366 </param>
|
|
|
367 <when value="basic">
|
|
|
368 <!-- Do nothing here -->
|
|
|
369 </when>
|
|
|
370 <when value="advanced">
|
|
|
371 <param name="filtered_records_merge_type" type="select" label="How should we deal with records seen at the same site in the VCF, but with different FILTER fields?" help="-filteredRecordsMergeType,--filteredrecordsmergetype &lt;filteredrecordsmergetype&gt;" >
|
|
|
372 <option value="KEEP_IF_ANY_UNFILTERED" selected="true"/>
|
|
|
373 <option value="KEEP_IF_ALL_UNFILTERED" />
|
|
|
374 </param>
|
|
|
375
|
|
|
376 <param name="print_complex_merges" checked="false" type="boolean" truevalue="--printComplexMerges" falsevalue="" label="Print out interesting sites requiring complex compatibility merging" help="-printComplexMerges,--printComplexMerges" />
|
|
|
377 <param name="filtered_are_uncalled" checked="false" type="boolean" truevalue="--filteredAreUncalled" falsevalue="" label="If true, then filtered VCFs are treated as uncalled, so that filtered set annotation don't appear in the combined VCF" help="-filteredAreUncalled,--filteredAreUncalled" />
|
|
|
378 <param name="minimal_vcf" checked="false" type="boolean" truevalue="--minimalVCF" falsevalue="" label="If true, then the output VCF will contain no INFO or genotype INFO field" help="-minimalVCF,--minimalVCF" />
|
|
|
379
|
|
|
380 <param name="set_key" type="text" value="" label="Key, by default set, in the INFO key=value tag emitted describing which set the combined VCF record came from." help="-setKey,--setKey &lt;setKey&gt;"/>
|
|
|
381 <param name="assume_identical_samples" checked="false" type="boolean" truevalue="--assumeIdenticalSamples" falsevalue="" label="If true, assume input VCFs have identical sample sets and disjoint calls so that one can simply perform a merge sort to combine the VCFs into one, drastically reducing the runtime." help="-assumeIdenticalSamples,--assumeIdenticalSamples" />
|
|
|
382 <param name="minimum_n" type="integer" value="1" label="Combine variants and output site only if variant is present in at least N input files." help="-minN,--minimumN &lt;minimumN&gt;"/>
|
|
|
383
|
|
|
384 </when>
|
|
|
385 </conditional>
|
|
|
386
|
|
|
387
|
|
|
388 </inputs>
|
|
|
389 <outputs>
|
|
|
390 <data format="vcf" name="output_variants" label="${tool.name} on ${on_string} (variants)" />
|
|
|
391 <data format="txt" name="output_log" label="${tool.name} on ${on_string} (log)" />
|
|
|
392 </outputs>
|
|
|
393 <tests>
|
|
|
394 <test>
|
|
|
395 <param name="reference_source_selector" value="history" />
|
|
|
396 <param name="ref_file" value="phiX.fasta" ftype="fasta" />
|
|
|
397 <param name="input_variant" value="gatk/gatk_variant_annotator/gatk_variant_annotator_out_1.vcf" ftype="vcf" />
|
|
|
398 <param name="input_variant_name" value="from_variant_annotator" />
|
|
|
399 <param name="genotype_merge_option" value="PRIORITIZE" />
|
|
|
400 <param name="gatk_param_type_selector" value="basic" />
|
|
|
401 <param name="analysis_param_type_selector" value="basic" />
|
|
|
402 <output name="output_variants" file="gatk/gatk_variant_combine/gatk_variant_combine_out_1.vcf" lines_diff="4" />
|
|
|
403 <output name="output_log" file="gatk/gatk_variant_combine/gatk_variant_combine_out_1.log.contains" compare="contains" />
|
|
|
404 </test>
|
|
|
405 </tests>
|
|
|
406 <help>
|
|
|
407 **What it does**
|
|
|
408
|
|
|
409 Combines VCF records from different sources; supports both full merges and set unions. Merge: combines multiple records into a single one; if sample names overlap then they are uniquified. Union: assumes each rod represents the same set of samples (although this is not enforced); using the priority list (if provided), emits a single record instance at every position represented in the rods.
|
|
|
410
|
|
|
411 For more information on using the CombineVariants module, see this `tool specific page <http://www.broadinstitute.org/gatk/gatkdocs/org_broadinstitute_sting_gatk_walkers_variantutils_CombineVariants.html>`_.
|
|
|
412
|
|
|
413 To learn about best practices for variant detection using GATK, see this `overview <http://www.broadinstitute.org/gatk/guide/topic?name=best-practices>`_.
|
|
|
414
|
|
|
415 If you encounter errors, please view the `GATK FAQ <http://www.broadinstitute.org/gatk/guide/topic?name=faqs>`_.
|
|
|
416
|
|
|
417 ------
|
|
|
418
|
|
|
419 **Inputs**
|
|
|
420
|
|
|
421 GenomeAnalysisTK: CombineVariants accepts variant files as input.
|
|
|
422
|
|
|
423 ------
|
|
|
424
|
|
|
425 **Outputs**
|
|
|
426
|
|
|
427 The output is a combined vcf file.
|
|
|
428
|
|
|
429
|
|
|
430 Go `here <http://www.broadinstitute.org/gatk/guide/topic?name=intro>`_ for details on GATK file formats.
|
|
|
431
|
|
|
432 -------
|
|
|
433
|
|
|
434 **Settings**::
|
|
|
435
|
|
|
436 out File to which variants should be written
|
|
|
437 genotypemergeoption How should we merge genotype records for samples shared across the ROD files? (UNIQUIFY|PRIORITIZE|UNSORTED|REQUIRE_UNIQUE)
|
|
|
438 filteredrecordsmergetype How should we deal with records seen at the same site in the VCF, but with different FILTER fields? KEEP_IF_ANY_UNFILTERED PASSes the record if any record is unfiltered, KEEP_IF_ALL_UNFILTERED requires all records to be unfiltered (KEEP_IF_ANY_UNFILTERED|KEEP_IF_ALL_UNFILTERED)
|
|
|
439 rod_priority_list When taking the union of variants containing genotypes: a comma-separated string describing the priority ordering for the genotypes as far as which record gets emitted; a complete priority list MUST be provided
|
|
|
440 printComplexMerges Print out interesting sites requiring complex compatibility merging
|
|
|
441 filteredAreUncalled If true, then filtered VCFs are treated as uncalled, so that filtered set annotation don't appear in the combined VCF
|
|
|
442 minimalVCF If true, then the output VCF will contain no INFO or genotype INFO field
|
|
|
443 setKey Key, by default set, in the INFO key=value tag emitted describing which set the combined VCF record came from. Set to null if you don't want the set field emitted.
|
|
|
444 assumeIdenticalSamples If true, assume input VCFs have identical sample sets and disjoint calls so that one can simply perform a merge sort to combine the VCFs into one, drastically reducing the runtime.
|
|
|
445 minimumN Combine variants and output site only if variant is present in at least N input files.
|
|
|
446
|
|
|
447 ------
|
|
|
448
|
|
|
449 **Citation**
|
|
|
450
|
|
|
451 For the underlying tool, please cite `DePristo MA, Banks E, Poplin R, Garimella KV, Maguire JR, Hartl C, Philippakis AA, del Angel G, Rivas MA, Hanna M, McKenna A, Fennell TJ, Kernytsky AM, Sivachenko AY, Cibulskis K, Gabriel SB, Altshuler D, Daly MJ. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat Genet. 2011 May;43(5):491-8. <http://www.ncbi.nlm.nih.gov/pubmed/21478889>`_
|
|
|
452
|
|
|
453 Please also site `McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K, Altshuler D, Gabriel S, Daly M, DePristo MA (2010). The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 20:1297-303. Epub 2010 Jul 19. <http://www.ncbi.nlm.nih.gov/pubmed/20644199>`_
|
|
|
454
|
|
|
455 If you use this tool in Galaxy, please cite `Blankenberg D, Von Kuster G, Coraor N, Ananda G, Lazarus R, Mangan M, Nekrutenko A, Taylor J. Galaxy: a web-based genome analysis tool for experimentalists. Curr Protoc Mol Biol. 2010 Jan;Chapter 19:Unit 19.10.1-21. <http://www.ncbi.nlm.nih.gov/pubmed/20069535>`_
|
|
|
456
|
|
|
457 </help>
|
|
|
458 </tool>
|
