|
0
|
1 <tool id="gatk2_indel_realigner" name="Indel Realigner" version="0.0.4">
|
|
|
2 <description>- perform local realignment</description>
|
|
|
3 <requirements>
|
|
|
4 <requirement type="package" version="2.2">gatk</requirement>
|
|
|
5 <requirement type="package" version="0.1.18">samtools</requirement>
|
|
|
6 </requirements>
|
|
|
7 <command interpreter="python">gatk2_wrapper.py
|
|
|
8 --max_jvm_heap_fraction "1"
|
|
|
9 --stdout "${output_log}"
|
|
|
10 -d "-I" "${reference_source.input_bam}" "${reference_source.input_bam.ext}" "gatk_input"
|
|
|
11 #if str( $reference_source.input_bam.metadata.bam_index ) != "None":
|
|
|
12 -d "" "${reference_source.input_bam.metadata.bam_index}" "bam_index" "gatk_input" ##hardcode galaxy ext type as bam_index
|
|
|
13 #end if
|
|
|
14 -p 'java
|
|
|
15 -jar "\$GATK2_PATH/GenomeAnalysisTK.jar"
|
|
|
16 -T "IndelRealigner"
|
|
|
17 -o "${output_bam}"
|
|
|
18 \$GATK2_SITE_OPTIONS
|
|
|
19 ##-et "NO_ET" -K "/data/galaxy/appList/GenomeAnalysisTK-2.0-36-gf5c1c1a/gatk2_key_file" ##ET no phone home
|
|
|
20 ##--num_threads 4 ##hard coded, for now
|
|
|
21 ##-log "${output_log}" ##don't use this to log to file, instead directly capture stdout
|
|
|
22 #if $reference_source.reference_source_selector != "history":
|
|
|
23 -R "${reference_source.ref_file.fields.path}"
|
|
|
24 #end if
|
|
|
25 -LOD "${lod_threshold}"
|
|
|
26 ${knowns_only}
|
|
|
27 '
|
|
|
28
|
|
|
29 #set $rod_binding_names = dict()
|
|
|
30 #for $rod_binding in $rod_bind:
|
|
|
31 #if str( $rod_binding.rod_bind_type.rod_bind_type_selector ) == 'custom':
|
|
|
32 #set $rod_bind_name = $rod_binding.rod_bind_type.custom_rod_name
|
|
|
33 #else
|
|
|
34 #set $rod_bind_name = $rod_binding.rod_bind_type.rod_bind_type_selector
|
|
|
35 #end if
|
|
|
36 #set $rod_binding_names[$rod_bind_name] = $rod_binding_names.get( $rod_bind_name, -1 ) + 1
|
|
|
37 -d "-known:${rod_bind_name},%(file_type)s" "${rod_binding.rod_bind_type.input_rod}" "${rod_binding.rod_bind_type.input_rod.ext}" "input_${rod_bind_name}_${rod_binding_names[$rod_bind_name]}"
|
|
|
38 #end for
|
|
|
39
|
|
|
40 ##start standard gatk options
|
|
|
41 #if $gatk_param_type.gatk_param_type_selector == "advanced":
|
|
|
42 #for $pedigree in $gatk_param_type.pedigree:
|
|
|
43 -p '--pedigree "${pedigree.pedigree_file}"'
|
|
|
44 #end for
|
|
|
45 #for $pedigree_string in $gatk_param_type.pedigree_string_repeat:
|
|
|
46 -p '--pedigreeString "${pedigree_string.pedigree_string}"'
|
|
|
47 #end for
|
|
|
48 -p '--pedigreeValidationType "${gatk_param_type.pedigree_validation_type}"'
|
|
|
49 #for $read_filter in $gatk_param_type.read_filter:
|
|
|
50 -p '--read_filter "${read_filter.read_filter_type.read_filter_type_selector}"
|
|
|
51 ###raise Exception( str( dir( $read_filter ) ) )
|
|
|
52 #for $name, $param in $read_filter.read_filter_type.iteritems():
|
|
|
53 #if $name not in [ "__current_case__", "read_filter_type_selector" ]:
|
|
|
54 #if hasattr( $param.input, 'truevalue' ):
|
|
|
55 ${param}
|
|
|
56 #else:
|
|
|
57 --${name} "${param}"
|
|
|
58 #end if
|
|
|
59 #end if
|
|
|
60 #end for
|
|
|
61 '
|
|
|
62 #end for
|
|
|
63 #for $interval_count, $input_intervals in enumerate( $gatk_param_type.input_interval_repeat ):
|
|
|
64 -d "--intervals" "${input_intervals.input_intervals}" "${input_intervals.input_intervals.ext}" "input_intervals_${interval_count}"
|
|
|
65 #end for
|
|
|
66
|
|
|
67 #for $interval_count, $input_intervals in enumerate( $gatk_param_type.input_exclude_interval_repeat ):
|
|
|
68 -d "--excludeIntervals" "${input_intervals.input_exclude_intervals}" "${input_intervals.input_exclude_intervals.ext}" "input_exlude_intervals_${interval_count}"
|
|
|
69 #end for
|
|
|
70
|
|
|
71 -p '--interval_set_rule "${gatk_param_type.interval_set_rule}"'
|
|
|
72
|
|
|
73 -p '--downsampling_type "${gatk_param_type.downsampling_type.downsampling_type_selector}"'
|
|
|
74 #if str( $gatk_param_type.downsampling_type.downsampling_type_selector ) != "NONE":
|
|
|
75 -p '--${gatk_param_type.downsampling_type.downsample_to_type.downsample_to_type_selector} "${gatk_param_type.downsampling_type.downsample_to_type.downsample_to_value}"'
|
|
|
76 #end if
|
|
|
77 -p '
|
|
|
78 --baq "${gatk_param_type.baq}"
|
|
|
79 --baqGapOpenPenalty "${gatk_param_type.baq_gap_open_penalty}"
|
|
|
80 ${gatk_param_type.use_original_qualities}
|
|
|
81 --defaultBaseQualities "${gatk_param_type.default_base_qualities}"
|
|
|
82 --validation_strictness "${gatk_param_type.validation_strictness}"
|
|
|
83 --interval_merging "${gatk_param_type.interval_merging}"
|
|
|
84 ${gatk_param_type.disable_experimental_low_memory_sharding}
|
|
|
85 ${gatk_param_type.non_deterministic_random_seed}
|
|
|
86 '
|
|
|
87 #for $rg_black_list_count, $rg_black_list in enumerate( $gatk_param_type.read_group_black_list_repeat ):
|
|
|
88 #if $rg_black_list.read_group_black_list_type.read_group_black_list_type_selector == "file":
|
|
|
89 -d "--read_group_black_list" "${rg_black_list.read_group_black_list_type.read_group_black_list}" "txt" "input_read_group_black_list_${rg_black_list_count}"
|
|
|
90 #else
|
|
|
91 -p '--read_group_black_list "${rg_black_list.read_group_black_list_type.read_group_black_list}"'
|
|
|
92 #end if
|
|
|
93 #end for
|
|
|
94 #end if
|
|
|
95 #if $reference_source.reference_source_selector == "history":
|
|
|
96 -d "-R" "${reference_source.ref_file}" "${reference_source.ref_file.ext}" "gatk_input"
|
|
|
97 #end if
|
|
|
98 ##end standard gatk options
|
|
|
99 ##start analysis specific options
|
|
|
100 -d "-targetIntervals" "${target_intervals}" "${target_intervals.ext}" "gatk_target_intervals"
|
|
|
101 -p '
|
|
|
102 --disable_bam_indexing
|
|
|
103 '
|
|
|
104 #if $analysis_param_type.analysis_param_type_selector == "advanced":
|
|
|
105 -p '
|
|
|
106 --entropyThreshold "${analysis_param_type.entropy_threshold}"
|
|
|
107 ${analysis_param_type.simplify_bam}
|
|
|
108 --consensusDeterminationModel "${analysis_param_type.consensus_determination_model}"
|
|
|
109 --maxIsizeForMovement "${analysis_param_type.max_insert_size_for_movement}"
|
|
|
110 --maxPositionalMoveAllowed "${analysis_param_type.max_positional_move_allowed}"
|
|
|
111 --maxConsensuses "${analysis_param_type.max_consensuses}"
|
|
|
112 --maxReadsForConsensuses "${analysis_param_type.max_reads_for_consensuses}"
|
|
|
113 --maxReadsForRealignment "${analysis_param_type.max_reads_for_realignment}"
|
|
|
114 ${analysis_param_type.no_original_alignment_tags}
|
|
|
115 '
|
|
|
116 #end if
|
|
|
117 </command>
|
|
|
118 <inputs>
|
|
|
119
|
|
|
120 <conditional name="reference_source">
|
|
|
121 <param name="reference_source_selector" type="select" label="Choose the source for the reference list">
|
|
|
122 <option value="cached">Locally cached</option>
|
|
|
123 <option value="history">History</option>
|
|
|
124 </param>
|
|
|
125 <when value="cached">
|
|
|
126 <param name="input_bam" type="data" format="bam" label="BAM file" help="-I,--input_file &lt;input_file&gt;">
|
|
|
127 <validator type="unspecified_build" />
|
|
|
128 <validator type="dataset_metadata_in_data_table" table_name="gatk2_picard_indexes" metadata_name="dbkey" metadata_column="dbkey" message="Sequences are not currently available for the specified build." /> <!-- fixme!!! this needs to be a select -->
|
|
|
129 </param>
|
|
|
130 <param name="ref_file" type="select" label="Using reference genome" help="-R,--reference_sequence &lt;reference_sequence&gt;" >
|
|
|
131 <options from_data_table="gatk2_picard_indexes">
|
|
|
132 <filter type="data_meta" key="dbkey" ref="input_bam" column="dbkey"/>
|
|
|
133 </options>
|
|
|
134 <validator type="no_options" message="A built-in reference genome is not available for the build associated with the selected input file"/>
|
|
|
135 </param>
|
|
|
136 </when>
|
|
|
137 <when value="history">
|
|
|
138 <param name="input_bam" type="data" format="bam" label="BAM file" help="-I,--input_file &lt;input_file&gt;" />
|
|
|
139 <param name="ref_file" type="data" format="fasta" label="Using reference file" help="-R,--reference_sequence &lt;reference_sequence&gt;">
|
|
|
140 <options>
|
|
|
141 <filter type="data_meta" key="dbkey" ref="input_bam" />
|
|
|
142 </options>
|
|
|
143 </param>
|
|
|
144 </when>
|
|
|
145 </conditional>
|
|
|
146 <param name="target_intervals" type="data" format="gatk_interval,bed,picard_interval_list" label="Restrict realignment to provided intervals" help="-targetIntervals,--targetIntervals &lt;targetIntervals&gt;" />
|
|
|
147 <repeat name="rod_bind" title="Binding for reference-ordered data" help="-known,--knownAlleles &lt;knownAlleles&gt;">
|
|
|
148 <conditional name="rod_bind_type">
|
|
|
149 <param name="rod_bind_type_selector" type="select" label="Binding Type">
|
|
|
150 <option value="dbsnp" selected="True">dbSNP</option>
|
|
|
151 <option value="snps">SNPs</option>
|
|
|
152 <option value="indels">INDELs</option>
|
|
|
153 <option value="custom">Custom</option>
|
|
|
154 </param>
|
|
|
155 <when value="dbsnp">
|
|
|
156 <param name="input_rod" type="data" format="vcf" label="ROD file" />
|
|
|
157 </when>
|
|
|
158 <when value="snps">
|
|
|
159 <param name="input_rod" type="data" format="vcf" label="ROD file" />
|
|
|
160 </when>
|
|
|
161 <when value="indels">
|
|
|
162 <param name="input_rod" type="data" format="vcf" label="ROD file" />
|
|
|
163 </when>
|
|
|
164 <when value="custom">
|
|
|
165 <param name="custom_rod_name" type="text" value="Unknown" label="ROD Name"/>
|
|
|
166 <param name="input_rod" type="data" format="vcf" label="ROD file" />
|
|
|
167 </when>
|
|
|
168 </conditional>
|
|
|
169 </repeat>
|
|
|
170 <param name="lod_threshold" type="float" value="5.0" label="LOD threshold above which the realigner will proceed to realign" help="-LOD,--LODThresholdForCleaning &lt;LODThresholdForCleaning&gt;" />
|
|
|
171 <param name="knowns_only" type="boolean" checked="False" truevalue="-knownsOnly" falsevalue="" label="Use only known indels provided as RODs" help="-knownsOnly"/>
|
|
|
172
|
|
|
173 <conditional name="gatk_param_type">
|
|
|
174 <param name="gatk_param_type_selector" type="select" label="Basic or Advanced GATK options">
|
|
|
175 <option value="basic" selected="True">Basic</option>
|
|
|
176 <option value="advanced">Advanced</option>
|
|
|
177 </param>
|
|
|
178 <when value="basic">
|
|
|
179 <!-- Do nothing here -->
|
|
|
180 </when>
|
|
|
181 <when value="advanced">
|
|
|
182 <repeat name="pedigree" title="Pedigree file" help="-ped,--pedigree &lt;pedigree&gt;">
|
|
|
183 <param name="pedigree_file" type="data" format="txt" label="Pedigree files for samples"/>
|
|
|
184 </repeat>
|
|
|
185 <repeat name="pedigree_string_repeat" title="Pedigree string" help="-pedString,--pedigreeString &lt;pedigreeString&gt;">
|
|
|
186 <param name="pedigree_string" type="text" value="" label="Pedigree string for samples"/>
|
|
|
187 </repeat>
|
|
|
188 <param name="pedigree_validation_type" type="select" label="How strict should we be in validating the pedigree information" help="-pedValidationType,--pedigreeValidationType &lt;pedigreeValidationType&gt;">
|
|
|
189 <option value="STRICT" selected="True">STRICT</option>
|
|
|
190 <option value="SILENT">SILENT</option>
|
|
|
191 </param>
|
|
|
192 <repeat name="read_filter" title="Read Filter" help="-rf,--read_filter &lt;read_filter&gt;">
|
|
|
193 <conditional name="read_filter_type">
|
|
|
194 <param name="read_filter_type_selector" type="select" label="Read Filter Type">
|
|
|
195 <option value="BadCigar">BadCigar</option>
|
|
|
196 <option value="BadMate">BadMate</option>
|
|
|
197 <option value="DuplicateRead">DuplicateRead</option>
|
|
|
198 <option value="FailsVendorQualityCheck">FailsVendorQualityCheck</option>
|
|
|
199 <option value="MalformedRead">MalformedRead</option>
|
|
|
200 <option value="MappingQuality">MappingQuality</option>
|
|
|
201 <option value="MappingQualityUnavailable">MappingQualityUnavailable</option>
|
|
|
202 <option value="MappingQualityZero">MappingQualityZero</option>
|
|
|
203 <option value="MateSameStrand">MateSameStrand</option>
|
|
|
204 <option value="MaxInsertSize">MaxInsertSize</option>
|
|
|
205 <option value="MaxReadLength" selected="True">MaxReadLength</option>
|
|
|
206 <option value="MissingReadGroup">MissingReadGroup</option>
|
|
|
207 <option value="NoOriginalQualityScores">NoOriginalQualityScores</option>
|
|
|
208 <option value="NotPrimaryAlignment">NotPrimaryAlignment</option>
|
|
|
209 <option value="Platform454">Platform454</option>
|
|
|
210 <option value="Platform">Platform</option>
|
|
|
211 <option value="PlatformUnit">PlatformUnit</option>
|
|
|
212 <option value="ReadGroupBlackList">ReadGroupBlackList</option>
|
|
|
213 <option value="ReadName">ReadName</option>
|
|
|
214 <option value="ReadStrand">ReadStrand</option>
|
|
|
215 <option value="ReassignMappingQuality">ReassignMappingQuality</option>
|
|
|
216 <option value="Sample">Sample</option>
|
|
|
217 <option value="SingleReadGroup">SingleReadGroup</option>
|
|
|
218 <option value="UnmappedRead">UnmappedRead</option>
|
|
|
219 </param>
|
|
|
220 <when value="BadCigar">
|
|
|
221 <!-- no extra options -->
|
|
|
222 </when>
|
|
|
223 <when value="BadMate">
|
|
|
224 <!-- no extra options -->
|
|
|
225 </when>
|
|
|
226 <when value="DuplicateRead">
|
|
|
227 <!-- no extra options -->
|
|
|
228 </when>
|
|
|
229 <when value="FailsVendorQualityCheck">
|
|
|
230 <!-- no extra options -->
|
|
|
231 </when>
|
|
|
232 <when value="MalformedRead">
|
|
|
233 <!-- no extra options -->
|
|
|
234 </when>
|
|
|
235 <when value="MappingQuality">
|
|
|
236 <param name="min_mapping_quality_score" type="integer" value="10" label="Minimum read mapping quality required to consider a read for calling"/>
|
|
|
237 </when>
|
|
|
238 <when value="MappingQualityUnavailable">
|
|
|
239 <!-- no extra options -->
|
|
|
240 </when>
|
|
|
241 <when value="MappingQualityZero">
|
|
|
242 <!-- no extra options -->
|
|
|
243 </when>
|
|
|
244 <when value="MateSameStrand">
|
|
|
245 <!-- no extra options -->
|
|
|
246 </when>
|
|
|
247 <when value="MaxInsertSize">
|
|
|
248 <param name="maxInsertSize" type="integer" value="1000000" label="Discard reads with insert size greater than the specified value"/>
|
|
|
249 </when>
|
|
|
250 <when value="MaxReadLength">
|
|
|
251 <param name="maxReadLength" type="integer" value="76" label="Max Read Length"/>
|
|
|
252 </when>
|
|
|
253 <when value="MissingReadGroup">
|
|
|
254 <!-- no extra options -->
|
|
|
255 </when>
|
|
|
256 <when value="NoOriginalQualityScores">
|
|
|
257 <!-- no extra options -->
|
|
|
258 </when>
|
|
|
259 <when value="NotPrimaryAlignment">
|
|
|
260 <!-- no extra options -->
|
|
|
261 </when>
|
|
|
262 <when value="Platform454">
|
|
|
263 <!-- no extra options -->
|
|
|
264 </when>
|
|
|
265 <when value="Platform">
|
|
|
266 <param name="PLFilterName" type="text" value="" label="Discard reads with RG:PL attribute containing this string"/>
|
|
|
267 </when>
|
|
|
268 <when value="PlatformUnit">
|
|
|
269 <!-- no extra options -->
|
|
|
270 </when>
|
|
|
271 <when value="ReadGroupBlackList">
|
|
|
272 <!-- no extra options -->
|
|
|
273 </when>
|
|
|
274 <when value="ReadName">
|
|
|
275 <param name="readName" type="text" value="" label="Filter out all reads except those with this read name"/>
|
|
|
276 </when>
|
|
|
277 <when value="ReadStrand">
|
|
|
278 <param name="filterPositive" type="boolean" truevalue="--filterPositive" falsevalue="" label="Discard reads on the forward strand"/>
|
|
|
279 </when>
|
|
|
280 <when value="ReassignMappingQuality">
|
|
|
281 <param name="default_mapping_quality" type="integer" value="60" label="Default read mapping quality to assign to all reads"/>
|
|
|
282 </when>
|
|
|
283 <when value="Sample">
|
|
|
284 <param name="sample_to_keep" type="text" value="" label="The name of the sample(s) to keep, filtering out all others"/>
|
|
|
285 </when>
|
|
|
286 <when value="SingleReadGroup">
|
|
|
287 <param name="read_group_to_keep" type="integer" value="76" label="The name of the read group to keep, filtering out all others"/>
|
|
|
288 </when>
|
|
|
289 <when value="UnmappedRead">
|
|
|
290 <!-- no extra options -->
|
|
|
291 </when>
|
|
|
292 </conditional>
|
|
|
293 </repeat>
|
|
|
294 <repeat name="input_interval_repeat" title="Operate on Genomic intervals" help="-L,--intervals &lt;intervals&gt;">
|
|
|
295 <param name="input_intervals" type="data" format="bed,gatk_interval,picard_interval_list,vcf" label="Genomic intervals" />
|
|
|
296 </repeat>
|
|
|
297 <repeat name="input_exclude_interval_repeat" title="Exclude Genomic intervals" help="-XL,--excludeIntervals &lt;excludeIntervals&gt;">
|
|
|
298 <param name="input_exclude_intervals" type="data" format="bed,gatk_interval,picard_interval_list,vcf" label="Genomic intervals" />
|
|
|
299 </repeat>
|
|
|
300
|
|
|
301 <param name="interval_set_rule" type="select" label="Interval set rule" help="-isr,--interval_set_rule &lt;interval_set_rule&gt;">
|
|
|
302 <option value="UNION" selected="True">UNION</option>
|
|
|
303 <option value="INTERSECTION">INTERSECTION</option>
|
|
|
304 </param>
|
|
|
305
|
|
|
306 <conditional name="downsampling_type">
|
|
|
307 <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;">
|
|
|
308 <option value="NONE" selected="True">NONE</option>
|
|
|
309 <option value="ALL_READS">ALL_READS</option>
|
|
|
310 <option value="BY_SAMPLE">BY_SAMPLE</option>
|
|
|
311 </param>
|
|
|
312 <when value="NONE">
|
|
|
313 <!-- no more options here -->
|
|
|
314 </when>
|
|
|
315 <when value="ALL_READS">
|
|
|
316 <conditional name="downsample_to_type">
|
|
|
317 <param name="downsample_to_type_selector" type="select" label="Downsample method">
|
|
|
318 <option value="downsample_to_fraction" selected="True">Downsample by Fraction</option>
|
|
|
319 <option value="downsample_to_coverage">Downsample by Coverage</option>
|
|
|
320 </param>
|
|
|
321 <when value="downsample_to_fraction">
|
|
|
322 <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;"/>
|
|
|
323 </when>
|
|
|
324 <when value="downsample_to_coverage">
|
|
|
325 <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;"/>
|
|
|
326 </when>
|
|
|
327 </conditional>
|
|
|
328 </when>
|
|
|
329 <when value="BY_SAMPLE">
|
|
|
330 <conditional name="downsample_to_type">
|
|
|
331 <param name="downsample_to_type_selector" type="select" label="Downsample method">
|
|
|
332 <option value="downsample_to_fraction" selected="True">Downsample by Fraction</option>
|
|
|
333 <option value="downsample_to_coverage">Downsample by Coverage</option>
|
|
|
334 </param>
|
|
|
335 <when value="downsample_to_fraction">
|
|
|
336 <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;"/>
|
|
|
337 </when>
|
|
|
338 <when value="downsample_to_coverage">
|
|
|
339 <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;"/>
|
|
|
340 </when>
|
|
|
341 </conditional>
|
|
|
342 </when>
|
|
|
343 </conditional>
|
|
|
344 <param name="baq" type="select" label="Type of BAQ calculation to apply in the engine" help="-baq,--baq &lt;baq&gt;">
|
|
|
345 <option value="OFF" selected="True">OFF</option>
|
|
|
346 <option value="CALCULATE_AS_NECESSARY">CALCULATE_AS_NECESSARY</option>
|
|
|
347 <option value="RECALCULATE">RECALCULATE</option>
|
|
|
348 </param>
|
|
|
349 <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;" />
|
|
|
350 <param name="use_original_qualities" type="boolean" truevalue="--useOriginalQualities" falsevalue="" label="Use the original base quality scores from the OQ tag" help="-OQ,--useOriginalQualities" />
|
|
|
351 <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;"/>
|
|
|
352 <param name="validation_strictness" type="select" label="How strict should we be with validation" help="-S,--validation_strictness &lt;validation_strictness&gt;">
|
|
|
353 <option value="STRICT" selected="True">STRICT</option>
|
|
|
354 <option value="LENIENT">LENIENT</option>
|
|
|
355 <option value="SILENT">SILENT</option>
|
|
|
356 <!-- <option value="DEFAULT_STRINGENCY">DEFAULT_STRINGENCY</option> listed in docs, but not valid value...-->
|
|
|
357 </param>
|
|
|
358 <param name="interval_merging" type="select" label="Interval merging rule" help="-im,--interval_merging &lt;interval_merging&gt;">
|
|
|
359 <option value="ALL" selected="True">ALL</option>
|
|
|
360 <option value="OVERLAPPING_ONLY">OVERLAPPING_ONLY</option>
|
|
|
361 </param>
|
|
|
362
|
|
|
363 <repeat name="read_group_black_list_repeat" title="Read group black list" help="-rgbl,--read_group_black_list &lt;read_group_black_list&gt;">
|
|
|
364 <conditional name="read_group_black_list_type">
|
|
|
365 <param name="read_group_black_list_type_selector" type="select" label="Type of reads read group black list">
|
|
|
366 <option value="file" selected="True">Filters in file</option>
|
|
|
367 <option value="text">Specify filters as a string</option>
|
|
|
368 </param>
|
|
|
369 <when value="file">
|
|
|
370 <param name="read_group_black_list" type="data" format="txt" label="Read group black list file" />
|
|
|
371 </when>
|
|
|
372 <when value="text">
|
|
|
373 <param name="read_group_black_list" type="text" value="tag:string" label="Read group black list tag:string" />
|
|
|
374 </when>
|
|
|
375 </conditional>
|
|
|
376 </repeat>
|
|
|
377
|
|
|
378 <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"/>
|
|
|
379 <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"/>
|
|
|
380
|
|
|
381 </when>
|
|
|
382 </conditional>
|
|
|
383
|
|
|
384 <conditional name="analysis_param_type">
|
|
|
385 <param name="analysis_param_type_selector" type="select" label="Basic or Advanced Analysis options">
|
|
|
386 <option value="basic" selected="True">Basic</option>
|
|
|
387 <option value="advanced">Advanced</option>
|
|
|
388 </param>
|
|
|
389 <when value="basic">
|
|
|
390 <!-- Do nothing here -->
|
|
|
391 </when>
|
|
|
392 <when value="advanced">
|
|
|
393
|
|
|
394 <param name="entropy_threshold" type="float" value="0.15" label="percentage of mismatching base quality scores at a position to be considered having high entropy" help="-entropy,--entropyThreshold &lt;entropyThreshold&gt;" />
|
|
|
395 <param name="simplify_bam" type="boolean" checked="False" truevalue="-simplifyBAM" falsevalue="" label="Simplify BAM" help="-simplifyBAM,--simplifyBAM"/>
|
|
|
396 <param name="consensus_determination_model" type="select" label="Consensus Determination Model" help="-model,--consensusDeterminationModel &lt;consensusDeterminationModel&gt;">
|
|
|
397 <option value="KNOWNS_ONLY">KNOWNS_ONLY</option>
|
|
|
398 <option value="USE_READS" selected="True">USE_READS</option>
|
|
|
399 <option value="USE_SW">USE_SW</option>
|
|
|
400 </param>
|
|
|
401 <param name="max_insert_size_for_movement" type="integer" value="3000" label="Maximum insert size of read pairs that we attempt to realign" help="-maxIsize,--maxIsizeForMovement &lt;maxIsizeForMovement&gt;" />
|
|
|
402 <param name="max_positional_move_allowed" type="integer" value="200" label="Maximum positional move in basepairs that a read can be adjusted during realignment" help="-maxPosMove,--maxPositionalMoveAllowed &lt;maxPositionalMoveAllowed&gt;" />
|
|
|
403 <param name="max_consensuses" type="integer" value="30" label="Max alternate consensuses to try" help="-maxConsensuses,--maxConsensuses &lt;maxConsensuses&gt;" />
|
|
|
404 <param name="max_reads_for_consensuses" type="integer" value="120" label="Max reads (chosen randomly) used for finding the potential alternate consensuses" help="-greedy,--maxReadsForConsensuses &lt;maxReadsForConsensuses&gt;" />
|
|
|
405 <param name="max_reads_for_realignment" type="integer" value="20000" label="Max reads allowed at an interval for realignment" help="-maxReads,--maxReadsForRealignment &lt;maxReadsForRealignment&gt;" />
|
|
|
406 <param name="no_original_alignment_tags" type="boolean" checked="False" truevalue="--noOriginalAlignmentTags" falsevalue="" label="Don't output the original cigar or alignment start tags for each realigned read in the output bam" help="-noTags,--noOriginalAlignmentTags"/>
|
|
|
407 </when>
|
|
|
408 </conditional>
|
|
|
409 </inputs>
|
|
|
410 <outputs>
|
|
|
411 <data format="bam" name="output_bam" label="${tool.name} on ${on_string} (BAM)" />
|
|
|
412 <data format="txt" name="output_log" label="${tool.name} on ${on_string} (log)" />
|
|
|
413 </outputs>
|
|
|
414 <tests>
|
|
|
415 <test>
|
|
|
416 <param name="reference_source_selector" value="history" />
|
|
|
417 <param name="ref_file" value="phiX.fasta" ftype="fasta" />
|
|
|
418 <param name="target_intervals" value="gatk/gatk_realigner_target_creator/gatk_realigner_target_creator_out_1.gatk_interval" ftype="gatk_interval" />
|
|
|
419 <param name="input_bam" value="gatk/fake_phiX_reads_1.bam" ftype="bam" />
|
|
|
420 <param name="rod_bind_type_selector" value="snps" />
|
|
|
421 <param name="input_rod" value="gatk/fake_phiX_variant_locations.vcf" ftype="vcf" />
|
|
|
422 <param name="lod_threshold" value="5.0" />
|
|
|
423 <param name="knowns_only" />
|
|
|
424 <param name="gatk_param_type_selector" value="basic" />
|
|
|
425 <param name="analysis_param_type_selector" value="advanced" />
|
|
|
426 <param name="entropy_threshold" value="0.15" />
|
|
|
427 <param name="simplify_bam" />
|
|
|
428 <param name="consensus_determination_model" value="USE_SW" />
|
|
|
429 <param name="max_insert_size_for_movement" value="3000" />
|
|
|
430 <param name="max_positional_move_allowed" value="200" />
|
|
|
431 <param name="max_consensuses" value="30" />
|
|
|
432 <param name="max_reads_for_consensuses" value="120" />
|
|
|
433 <param name="max_reads_for_realignment" value="20000" />
|
|
|
434 <param name="no_original_alignment_tags" />
|
|
|
435 <output name="output_bam" file="gatk/gatk_indel_realigner/gatk_indel_realigner_out_1.bam" ftype="bam" lines_diff="2" />
|
|
|
436 <output name="output_log" file="gatk/gatk_indel_realigner/gatk_indel_realigner_out_1.log.contains" compare="contains" />
|
|
|
437 </test>
|
|
|
438 </tests>
|
|
|
439 <help>
|
|
|
440 **What it does**
|
|
|
441
|
|
|
442 Performs local realignment of reads based on misalignments due to the presence of indels. Unlike most mappers, this walker uses the full alignment context to determine whether an appropriate alternate reference (i.e. indel) exists and updates SAMRecords accordingly.
|
|
|
443
|
|
|
444 For more information on local realignment around indels using the GATK, see this `tool specific page <http://www.broadinstitute.org/gatk/gatkdocs/org_broadinstitute_sting_gatk_walkers_indels_IndelRealigner.html>`_.
|
|
|
445
|
|
|
446 To learn about best practices for variant detection using GATK, see this `overview <http://www.broadinstitute.org/gatk/guide/topic?name=best-practices>`_.
|
|
|
447
|
|
|
448 If you encounter errors, please view the `GATK FAQ <http://www.broadinstitute.org/gatk/guide/topic?name=faqs>`_.
|
|
|
449
|
|
|
450 ------
|
|
|
451
|
|
|
452 **Inputs**
|
|
|
453
|
|
|
454 GenomeAnalysisTK: IndelRealigner accepts an aligned BAM and a list of intervals to realign as input files.
|
|
|
455
|
|
|
456
|
|
|
457 **Outputs**
|
|
|
458
|
|
|
459 The output is in the BAM format.
|
|
|
460
|
|
|
461
|
|
|
462 Go `here <http://www.broadinstitute.org/gatk/guide/topic?name=intro>`_ for details on GATK file formats.
|
|
|
463
|
|
|
464 -------
|
|
|
465
|
|
|
466 **Settings**::
|
|
|
467
|
|
|
468 targetIntervals intervals file output from RealignerTargetCreator
|
|
|
469 LODThresholdForCleaning LOD threshold above which the cleaner will clean
|
|
|
470 entropyThreshold percentage of mismatches at a locus to be considered having high entropy
|
|
|
471 out Output bam
|
|
|
472 bam_compression Compression level to use for writing BAM files
|
|
|
473 disable_bam_indexing Turn off on-the-fly creation of indices for output BAM files.
|
|
|
474 simplifyBAM If provided, output BAM files will be simplified to include just key reads for downstream variation discovery analyses (removing duplicates, PF-, non-primary reads), as well stripping all extended tags from the kept reads except the read group identifier
|
|
|
475 useOnlyKnownIndels Don't run 'Smith-Waterman' to generate alternate consenses; use only known indels provided as RODs for constructing the alternate references.
|
|
|
476 maxReadsInMemory max reads allowed to be kept in memory at a time by the SAMFileWriter. Keep it low to minimize memory consumption (but the tool may skip realignment on regions with too much coverage. If it is too low, it may generate errors during realignment); keep it high to maximize realignment (but make sure to give Java enough memory).
|
|
|
477 maxIsizeForMovement maximum insert size of read pairs that we attempt to realign
|
|
|
478 maxPositionalMoveAllowed maximum positional move in basepairs that a read can be adjusted during realignment
|
|
|
479 maxConsensuses max alternate consensuses to try (necessary to improve performance in deep coverage)
|
|
|
480 maxReadsForConsensuses max reads used for finding the alternate consensuses (necessary to improve performance in deep coverage)
|
|
|
481 maxReadsForRealignment max reads allowed at an interval for realignment; if this value is exceeded, realignment is not attempted and the reads are passed to the output file(s) as-is
|
|
|
482 noOriginalAlignmentTags Don't output the original cigar or alignment start tags for each realigned read in the output bam.
|
|
|
483
|
|
|
484 ------
|
|
|
485
|
|
|
486 **Citation**
|
|
|
487
|
|
|
488 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>`_
|
|
|
489
|
|
|
490 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>`_
|
|
|
491
|
|
|
492 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>`_
|
|
|
493
|
|
|
494 </help>
|
|
|
495 </tool>
|
