Mercurial > repos > jjohnson > gatk2
annotate reduce_reads.xml @ 11:5faf7ace8aee draft
Add HaplotypeCaller
| author | Jim Johnson <jj@umn.edu> |
|---|---|
| date | Thu, 08 Nov 2012 10:38:18 -0600 |
| parents | a14e79e7ac75 |
| children | d73c92a7b0ea |
| rev | line source |
|---|---|
| 6 | 1 <tool id="gatk2_reduce_reads" name="Reduce Reads" version="0.0.4"> |
| 2 <description>in BAM files</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 "ReduceReads" | |
| 17 -o "${output_bam}" | |
|
10
a14e79e7ac75
Comment out site options from reduce_reads.xml
Jim Johnson <jj@umn.edu>
parents:
7
diff
changeset
|
18 ## \$GATK2_SITE_OPTIONS |
| 6 | 19 ##-et "NO_ET" -K "/data/galaxy/appList/GenomeAnalysisTK-2.0-36-gf5c1c1a/gatk2_key_file" ##ET no phone home |
| 20 ##--num_threads 4 ##not supported yet | |
| 21 ##-log "${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 #if str($input_recal) != 'None': | |
| 26 --BQSR "${input_recal}" | |
| 27 #end if | |
| 28 --disable_bam_indexing | |
| 29 ' | |
| 30 ##start standard gatk options | |
| 31 #if $gatk_param_type.gatk_param_type_selector == "advanced": | |
| 32 #for $pedigree in $gatk_param_type.pedigree: | |
| 33 -p '--pedigree "${pedigree.pedigree_file}"' | |
| 34 #end for | |
| 35 #for $pedigree_string in $gatk_param_type.pedigree_string_repeat: | |
| 36 -p '--pedigreeString "${pedigree_string.pedigree_string}"' | |
| 37 #end for | |
| 38 -p '--pedigreeValidationType "${gatk_param_type.pedigree_validation_type}"' | |
| 39 #for $read_filter in $gatk_param_type.read_filter: | |
| 40 -p '--read_filter "${read_filter.read_filter_type.read_filter_type_selector}" | |
| 41 ###raise Exception( str( dir( $read_filter ) ) ) | |
| 42 #for $name, $param in $read_filter.read_filter_type.iteritems(): | |
| 43 #if $name not in [ "__current_case__", "read_filter_type_selector" ]: | |
| 44 #if hasattr( $param.input, 'truevalue' ): | |
| 45 ${param} | |
| 46 #else: | |
| 47 --${name} "${param}" | |
| 48 #end if | |
| 49 #end if | |
| 50 #end for | |
| 51 ' | |
| 52 #end for | |
| 53 #for $interval_count, $input_intervals in enumerate( $gatk_param_type.input_interval_repeat ): | |
| 54 -d "--intervals" "${input_intervals.input_intervals}" "${input_intervals.input_intervals.ext}" "input_intervals_${interval_count}" | |
| 55 #end for | |
| 56 | |
| 57 #for $interval_count, $input_intervals in enumerate( $gatk_param_type.input_exclude_interval_repeat ): | |
| 58 -d "--excludeIntervals" "${input_intervals.input_exclude_intervals}" "${input_intervals.input_exclude_intervals.ext}" "input_exlude_intervals_${interval_count}" | |
| 59 #end for | |
| 60 | |
| 61 -p '--interval_set_rule "${gatk_param_type.interval_set_rule}"' | |
| 62 | |
| 63 -p '--downsampling_type "${gatk_param_type.downsampling_type.downsampling_type_selector}"' | |
| 64 #if str( $gatk_param_type.downsampling_type.downsampling_type_selector ) != "NONE": | |
| 65 -p '--${gatk_param_type.downsampling_type.downsample_to_type.downsample_to_type_selector} "${gatk_param_type.downsampling_type.downsample_to_type.downsample_to_value}"' | |
| 66 #end if | |
| 67 -p ' | |
| 68 --baq "${gatk_param_type.baq}" | |
| 69 --baqGapOpenPenalty "${gatk_param_type.baq_gap_open_penalty}" | |
| 70 ${gatk_param_type.use_original_qualities} | |
| 71 --defaultBaseQualities "${gatk_param_type.default_base_qualities}" | |
| 72 --validation_strictness "${gatk_param_type.validation_strictness}" | |
| 73 --interval_merging "${gatk_param_type.interval_merging}" | |
| 74 ${gatk_param_type.disable_experimental_low_memory_sharding} | |
| 75 ${gatk_param_type.non_deterministic_random_seed} | |
| 76 ' | |
| 77 #for $rg_black_list_count, $rg_black_list in enumerate( $gatk_param_type.read_group_black_list_repeat ): | |
| 78 #if $rg_black_list.read_group_black_list_type.read_group_black_list_type_selector == "file": | |
| 79 -d "--read_group_black_list" "${rg_black_list.read_group_black_list_type.read_group_black_list}" "txt" "input_read_group_black_list_${rg_black_list_count}" | |
| 80 #else | |
| 81 -p '--read_group_black_list "${rg_black_list.read_group_black_list_type.read_group_black_list}"' | |
| 82 #end if | |
| 83 #end for | |
| 84 #end if | |
| 85 | |
| 86 #if str( $reference_source.reference_source_selector ) == "history": | |
| 87 -d "-R" "${reference_source.ref_file}" "${reference_source.ref_file.ext}" "gatk_input" | |
| 88 #end if | |
| 89 ##end standard gatk options | |
| 90 | |
| 91 ##start analysis specific options | |
| 92 #if $analysis_param_type.analysis_param_type_selector == "advanced": | |
| 93 -p ' | |
| 94 #if $analysis_param_type.context_size.__str__.strip() != '': | |
| 95 --context_size $analysis_param_type.context_size | |
| 96 #end if | |
| 97 #if $analysis_param_type.downsample_coverage.__str__.strip() != '': | |
| 98 --downsample_coverage $analysis_param_type.downsample_coverage | |
| 99 #end if | |
| 100 #if $analysis_param_type.minimum_del_proportion_to_trigger_variant.__str__.strip() != '': | |
| 101 --minimum_del_proportion_to_trigger_variant $analysis_param_type.minimum_del_proportion_to_trigger_variant | |
| 102 #end if | |
| 103 #if $analysis_param_type.minimum_mapping_quality.__str__.strip() != '': | |
| 104 --minimum_mapping_quality $analysis_param_type.minimum_mapping_quality | |
| 105 #end if | |
| 106 #if $analysis_param_type.minimum_tail_qualities.__str__.strip() != '': | |
| 107 --minimum_tail_qualities $analysis_param_type.minimum_tail_qualities | |
| 108 #end if | |
| 109 #if $analysis_param_type.minimum_base_quality_to_consider.__str__.strip() != '': | |
| 110 --minimum_base_quality_to_consider $analysis_param_type.minimum_base_quality_to_consider | |
| 111 #end if | |
| 112 #if $analysis_param_type.minimum_alt_proportion_to_trigger_variant.__str__.strip() != '': | |
| 113 --minimum_alt_proportion_to_trigger_variant $analysis_param_type.minimum_alt_proportion_to_trigger_variant | |
| 114 #end if | |
| 115 $analysis_param_type.allow_polyploid_reduction | |
| 116 $analysis_param_type.dont_compress_read_names | |
| 117 $analysis_param_type.dont_hardclip_low_qual_tails | |
| 118 $analysis_param_type.dont_simplify_reads | |
| 119 $analysis_param_type.dont_use_softclipped_bases | |
| 120 $analysis_param_type.hard_clip_to_interval | |
| 121 $analysis_param_type.dont_hardclip_adaptor_sequences | |
| 122 ' | |
| 123 #end if | |
| 124 </command> | |
| 125 <inputs> | |
| 11 | 126 <param name="input_recal" type="data" format="csv" optional="true" label="Covariates table recalibration file" help="-BQSR,--BQSR &lt;recal_file&gt;" > |
| 127 <help>The input covariates table file which enables on-the-fly base quality score recalibration. | |
| 128 Enables on-the-fly recalibrate of base qualities. The covariates tables are produced by the BaseQualityScoreRecalibrator tool. | |
| 129 Please be aware that one should only run recalibration with the covariates file created on the same input bam(s). | |
| 130 </help> | |
| 131 </param> | |
| 6 | 132 <conditional name="reference_source"> |
| 133 <param name="reference_source_selector" type="select" label="Choose the source for the reference list"> | |
| 134 <option value="cached">Locally cached</option> | |
| 135 <option value="history">History</option> | |
| 136 </param> | |
| 137 <when value="cached"> | |
| 138 <param name="input_bam" type="data" format="bam" label="BAM file" help="-I,--input_file &lt;input_file&gt;"> | |
| 139 <validator type="unspecified_build" /> | |
| 140 <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 --> | |
| 141 </param> | |
| 142 <param name="ref_file" type="select" label="Using reference genome" help="-R,--reference_sequence &lt;reference_sequence&gt;" > | |
| 143 <options from_data_table="gatk2_picard_indexes"> | |
| 144 <filter type="data_meta" key="dbkey" ref="input_bam" column="dbkey"/> | |
| 145 </options> | |
| 146 <validator type="no_options" message="A built-in reference genome is not available for the build associated with the selected input file"/> | |
| 147 </param> | |
| 148 </when> | |
| 149 <when value="history"> | |
| 150 <param name="input_bam" type="data" format="bam" label="BAM file" help="-I,--input_file &lt;input_file&gt;" /> | |
| 151 <param name="ref_file" type="data" format="fasta" label="Using reference file" help="-R,--reference_sequence &lt;reference_sequence&gt;"> | |
| 152 <options> | |
| 153 <filter type="data_meta" key="dbkey" ref="input_bam" /> | |
| 154 </options> | |
| 155 </param> | |
| 156 </when> | |
| 157 </conditional> | |
| 158 | |
| 159 <conditional name="gatk_param_type"> | |
| 160 <param name="gatk_param_type_selector" type="select" label="Basic or Advanced GATK options"> | |
| 161 <option value="basic" selected="True">Basic</option> | |
| 162 <option value="advanced">Advanced</option> | |
| 163 </param> | |
| 164 <when value="basic"> | |
| 165 <!-- Do nothing here --> | |
| 166 </when> | |
| 167 <when value="advanced"> | |
| 168 <repeat name="pedigree" title="Pedigree file" help="-ped,--pedigree &lt;pedigree&gt;"> | |
| 169 <param name="pedigree_file" type="data" format="txt" label="Pedigree files for samples"/> | |
| 170 </repeat> | |
| 171 <repeat name="pedigree_string_repeat" title="Pedigree string" help="-pedString,--pedigreeString &lt;pedigreeString&gt;"> | |
| 172 <param name="pedigree_string" type="text" value="" label="Pedigree string for samples"/> | |
| 173 </repeat> | |
| 174 <param name="pedigree_validation_type" type="select" label="How strict should we be in validating the pedigree information" help="-pedValidationType,--pedigreeValidationType &lt;pedigreeValidationType&gt;"> | |
| 175 <option value="STRICT" selected="True">STRICT</option> | |
| 176 <option value="SILENT">SILENT</option> | |
| 177 </param> | |
| 178 <repeat name="read_filter" title="Read Filter" help="-rf,--read_filter &lt;read_filter&gt;"> | |
| 179 <conditional name="read_filter_type"> | |
| 180 <param name="read_filter_type_selector" type="select" label="Read Filter Type"> | |
| 181 <option value="BadCigar">BadCigar</option> | |
| 182 <option value="BadMate">BadMate</option> | |
| 183 <option value="DuplicateRead">DuplicateRead</option> | |
| 184 <option value="FailsVendorQualityCheck">FailsVendorQualityCheck</option> | |
| 185 <option value="MalformedRead">MalformedRead</option> | |
| 186 <option value="MappingQuality">MappingQuality</option> | |
| 187 <option value="MappingQualityUnavailable">MappingQualityUnavailable</option> | |
| 188 <option value="MappingQualityZero">MappingQualityZero</option> | |
| 189 <option value="MateSameStrand">MateSameStrand</option> | |
| 190 <option value="MaxInsertSize">MaxInsertSize</option> | |
| 191 <option value="MaxReadLength" selected="True">MaxReadLength</option> | |
| 192 <option value="MissingReadGroup">MissingReadGroup</option> | |
| 193 <option value="NoOriginalQualityScores">NoOriginalQualityScores</option> | |
| 194 <option value="NotPrimaryAlignment">NotPrimaryAlignment</option> | |
| 195 <option value="Platform454">Platform454</option> | |
| 196 <option value="Platform">Platform</option> | |
| 197 <option value="PlatformUnit">PlatformUnit</option> | |
| 198 <option value="ReadGroupBlackList">ReadGroupBlackList</option> | |
| 199 <option value="ReadName">ReadName</option> | |
| 200 <option value="ReadStrand">ReadStrand</option> | |
| 201 <option value="ReassignMappingQuality">ReassignMappingQuality</option> | |
| 202 <option value="Sample">Sample</option> | |
| 203 <option value="SingleReadGroup">SingleReadGroup</option> | |
| 204 <option value="UnmappedRead">UnmappedRead</option> | |
| 205 </param> | |
| 206 <when value="BadCigar"> | |
| 207 <!-- no extra options --> | |
| 208 </when> | |
| 209 <when value="BadMate"> | |
| 210 <!-- no extra options --> | |
| 211 </when> | |
| 212 <when value="DuplicateRead"> | |
| 213 <!-- no extra options --> | |
| 214 </when> | |
| 215 <when value="FailsVendorQualityCheck"> | |
| 216 <!-- no extra options --> | |
| 217 </when> | |
| 218 <when value="MalformedRead"> | |
| 219 <!-- no extra options --> | |
| 220 </when> | |
| 221 <when value="MappingQuality"> | |
| 222 <param name="min_mapping_quality_score" type="integer" value="10" label="Minimum read mapping quality required to consider a read for calling"/> | |
| 223 </when> | |
| 224 <when value="MappingQualityUnavailable"> | |
| 225 <!-- no extra options --> | |
| 226 </when> | |
| 227 <when value="MappingQualityZero"> | |
| 228 <!-- no extra options --> | |
| 229 </when> | |
| 230 <when value="MateSameStrand"> | |
| 231 <!-- no extra options --> | |
| 232 </when> | |
| 233 <when value="MaxInsertSize"> | |
| 234 <param name="maxInsertSize" type="integer" value="1000000" label="Discard reads with insert size greater than the specified value"/> | |
| 235 </when> | |
| 236 <when value="MaxReadLength"> | |
| 237 <param name="maxReadLength" type="integer" value="76" label="Max Read Length"/> | |
| 238 </when> | |
| 239 <when value="MissingReadGroup"> | |
| 240 <!-- no extra options --> | |
| 241 </when> | |
| 242 <when value="NoOriginalQualityScores"> | |
| 243 <!-- no extra options --> | |
| 244 </when> | |
| 245 <when value="NotPrimaryAlignment"> | |
| 246 <!-- no extra options --> | |
| 247 </when> | |
| 248 <when value="Platform454"> | |
| 249 <!-- no extra options --> | |
| 250 </when> | |
| 251 <when value="Platform"> | |
| 252 <param name="PLFilterName" type="text" value="" label="Discard reads with RG:PL attribute containing this string"/> | |
| 253 </when> | |
| 254 <when value="PlatformUnit"> | |
| 255 <!-- no extra options --> | |
| 256 </when> | |
| 257 <when value="ReadGroupBlackList"> | |
| 258 <!-- no extra options --> | |
| 259 </when> | |
| 260 <when value="ReadName"> | |
| 261 <param name="readName" type="text" value="" label="Filter out all reads except those with this read name"/> | |
| 262 </when> | |
| 263 <when value="ReadStrand"> | |
| 264 <param name="filterPositive" type="boolean" truevalue="--filterPositive" falsevalue="" label="Discard reads on the forward strand"/> | |
| 265 </when> | |
| 266 <when value="ReassignMappingQuality"> | |
| 267 <param name="default_mapping_quality" type="integer" value="60" label="Default read mapping quality to assign to all reads"/> | |
| 268 </when> | |
| 269 <when value="Sample"> | |
| 270 <param name="sample_to_keep" type="text" value="" label="The name of the sample(s) to keep, filtering out all others"/> | |
| 271 </when> | |
| 272 <when value="SingleReadGroup"> | |
| 273 <param name="read_group_to_keep" type="integer" value="76" label="The name of the read group to keep, filtering out all others"/> | |
| 274 </when> | |
| 275 <when value="UnmappedRead"> | |
| 276 <!-- no extra options --> | |
| 277 </when> | |
| 278 </conditional> | |
| 279 </repeat> | |
| 280 <repeat name="input_interval_repeat" title="Operate on Genomic intervals" help="-L,--intervals &lt;intervals&gt;"> | |
| 281 <param name="input_intervals" type="data" format="bed,gatk_interval,picard_interval_list,vcf" label="Genomic intervals" /> | |
| 282 </repeat> | |
| 283 <repeat name="input_exclude_interval_repeat" title="Exclude Genomic intervals" help="-XL,--excludeIntervals &lt;excludeIntervals&gt;"> | |
| 284 <param name="input_exclude_intervals" type="data" format="bed,gatk_interval,picard_interval_list,vcf" label="Genomic intervals" /> | |
| 285 </repeat> | |
| 286 | |
| 287 <param name="interval_set_rule" type="select" label="Interval set rule" help="-isr,--interval_set_rule &lt;interval_set_rule&gt;"> | |
| 288 <option value="UNION" selected="True">UNION</option> | |
| 289 <option value="INTERSECTION">INTERSECTION</option> | |
| 290 </param> | |
| 291 | |
| 292 <conditional name="downsampling_type"> | |
| 293 <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;"> | |
| 294 <option value="NONE" selected="True">NONE</option> | |
| 295 <option value="ALL_READS">ALL_READS</option> | |
| 296 <option value="BY_SAMPLE">BY_SAMPLE</option> | |
| 297 </param> | |
| 298 <when value="NONE"> | |
| 299 <!-- no more options here --> | |
| 300 </when> | |
| 301 <when value="ALL_READS"> | |
| 302 <conditional name="downsample_to_type"> | |
| 303 <param name="downsample_to_type_selector" type="select" label="Downsample method"> | |
| 304 <option value="downsample_to_fraction" selected="True">Downsample by Fraction</option> | |
| 305 <option value="downsample_to_coverage">Downsample by Coverage</option> | |
| 306 </param> | |
| 307 <when value="downsample_to_fraction"> | |
| 308 <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;"/> | |
| 309 </when> | |
| 310 <when value="downsample_to_coverage"> | |
| 311 <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;"/> | |
| 312 </when> | |
| 313 </conditional> | |
| 314 </when> | |
| 315 <when value="BY_SAMPLE"> | |
| 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 </conditional> | |
| 330 <param name="baq" type="select" label="Type of BAQ calculation to apply in the engine" help="-baq,--baq &lt;baq&gt;"> | |
| 331 <option value="OFF" selected="True">OFF</option> | |
| 332 <option value="CALCULATE_AS_NECESSARY">CALCULATE_AS_NECESSARY</option> | |
| 333 <option value="RECALCULATE">RECALCULATE</option> | |
| 334 </param> | |
| 335 <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;" /> | |
| 336 <param name="use_original_qualities" type="boolean" truevalue="--useOriginalQualities" falsevalue="" label="Use the original base quality scores from the OQ tag" help="-OQ,--useOriginalQualities" /> | |
| 337 <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;"/> | |
| 338 <param name="validation_strictness" type="select" label="How strict should we be with validation" help="-S,--validation_strictness &lt;validation_strictness&gt;"> | |
| 339 <option value="STRICT" selected="True">STRICT</option> | |
| 340 <option value="LENIENT">LENIENT</option> | |
| 341 <option value="SILENT">SILENT</option> | |
| 342 <!-- <option value="DEFAULT_STRINGENCY">DEFAULT_STRINGENCY</option> listed in docs, but not valid value...--> | |
| 343 </param> | |
| 344 <param name="interval_merging" type="select" label="Interval merging rule" help="-im,--interval_merging &lt;interval_merging&gt;"> | |
| 345 <option value="ALL" selected="True">ALL</option> | |
| 346 <option value="OVERLAPPING_ONLY">OVERLAPPING_ONLY</option> | |
| 347 </param> | |
| 348 | |
| 349 <repeat name="read_group_black_list_repeat" title="Read group black list" help="-rgbl,--read_group_black_list &lt;read_group_black_list&gt;"> | |
| 350 <conditional name="read_group_black_list_type"> | |
| 351 <param name="read_group_black_list_type_selector" type="select" label="Type of reads read group black list"> | |
| 352 <option value="file" selected="True">Filters in file</option> | |
| 353 <option value="text">Specify filters as a string</option> | |
| 354 </param> | |
| 355 <when value="file"> | |
| 356 <param name="read_group_black_list" type="data" format="txt" label="Read group black list file" /> | |
| 357 </when> | |
| 358 <when value="text"> | |
| 359 <param name="read_group_black_list" type="text" value="tag:string" label="Read group black list tag:string" /> | |
| 360 </when> | |
| 361 </conditional> | |
| 362 </repeat> | |
| 363 | |
| 364 <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"/> | |
| 365 <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"/> | |
| 366 | |
| 367 </when> | |
| 368 </conditional> | |
| 369 | |
| 370 | |
| 371 <conditional name="analysis_param_type"> | |
| 372 <param name="analysis_param_type_selector" type="select" label="Basic or Advanced Analysis options"> | |
| 373 <option value="basic" selected="True">Basic</option> | |
| 374 <option value="advanced">Advanced</option> | |
| 375 </param> | |
| 376 <when value="basic"> | |
| 377 <!-- Do nothing here --> | |
| 378 </when> | |
| 379 <when value="advanced"> | |
| 380 <conditional name="default_read_group_type"> | |
| 381 <param name="default_read_group_type_selector" type="select" label="Set default Read Group" help="--default_read_group"> | |
| 382 <option value="default" selected="True">Don't Set</option> | |
| 383 <option value="set">Set</option> | |
| 384 </param> | |
| 385 <when value="default"> | |
| 386 <!-- do nothing here --> | |
| 387 </when> | |
| 388 <when value="set"> | |
| 389 <param name="default_read_group" type="text" value="Unknown" label="If a read has no read group then default to the provided String"/> | |
| 390 </when> | |
| 391 </conditional> | |
| 392 <param name="default_platform" type="select" label="Set default Platform" help="--default_platform"> | |
| 393 <option value="default" selected="True">Don't Set</option> | |
| 394 <option value="illumina">illumina</option> | |
| 395 <option value="454">454</option> | |
| 396 <option value="solid">solid</option> | |
| 397 </param> | |
| 398 <conditional name="force_read_group_type"> | |
| 399 <param name="force_read_group_type_selector" type="select" label="Force Read Group" help="--force_read_group"> | |
| 400 <option value="default" selected="True">Don't Force</option> | |
| 401 <option value="set">Force</option> | |
| 402 </param> | |
| 403 <when value="default"> | |
| 404 <!-- do nothing here --> | |
| 405 </when> | |
| 406 <when value="set"> | |
| 407 <param name="force_read_group" type="text" value="Unknown" label="If provided, the read group ID of EVERY read will be forced to be the provided String."/> | |
| 408 </when> | |
| 409 </conditional> | |
| 410 <param name="force_platform" type="select" label="Force Platform" help="--force_platform"> | |
| 411 <option value="default" selected="True">Don't Force</option> | |
| 412 <option value="illumina">illumina</option> | |
| 413 <option value="454">454</option> | |
| 414 <option value="solid">solid</option> | |
| 415 </param> | |
| 416 <param name="exception_if_no_tile" type="boolean" checked="False" truevalue="--exception_if_no_tile" falsevalue="" label="Throw an exception when no tile can be found" help="--exception_if_no_tile"/> | |
| 417 <conditional name="solid_options_type"> | |
| 418 <param name="solid_options_type_selector" type="select" label="Set SOLiD specific options"> | |
| 419 <option value="default" selected="True">Don't Set</option> | |
| 420 <option value="set">Set</option> | |
| 421 </param> | |
| 422 <when value="default"> | |
| 423 <!-- do nothing here --> | |
| 424 </when> | |
| 425 <when value="set"> | |
| 426 <param name="solid_recal_mode" type="select" label="How should we recalibrate solid bases in which the reference was inserted" help="-sMode,--solid_recal_mode &lt;solid_recal_mode&gt;"> | |
| 427 <option value="default" selected="True">Don't set</option> | |
| 428 <option value="DO_NOTHING">DO_NOTHING</option> | |
| 429 <option value="SET_Q_ZERO">SET_Q_ZERO</option> | |
| 430 <option value="SET_Q_ZERO_BASE_N">SET_Q_ZERO_BASE_N</option> | |
| 431 <option value="REMOVE_REF_BIAS">REMOVE_REF_BIAS</option> | |
| 432 </param> | |
| 433 <param name="solid_nocall_strategy" type="select" label="Behavior of the recalibrator when it encounters no calls" help="-solid_nocall_strategy,--solid_nocall_strategy &lt;solid_nocall_strategy&gt;"> | |
| 434 <option value="default" selected="True">Don't set</option> | |
| 435 <option value="THROW_EXCEPTION">THROW_EXCEPTION</option> | |
| 436 <option value="LEAVE_READ_UNRECALIBRATED">LEAVE_READ_UNRECALIBRATED</option> | |
| 437 <option value="PURGE_READ">PURGE_READ</option> | |
| 438 </param> | |
| 439 </when> | |
| 440 </conditional> | |
| 441 <param name="simplify_bam" type="boolean" checked="False" truevalue="-simplifyBAM" falsevalue="" label="Simplify BAM" help="-simplifyBAM,--simplifyBAM"/> | |
| 442 <param name="window_size_nqs" type="integer" value="5" label="Window size used by MinimumNQSCovariate" help="--window_size_nqs"/> | |
| 443 <param name="homopolymer_nback" type="integer" value="7" label="Number of previous bases to look at in HomopolymerCovariate" help="-nback,--homopolymer_nback &lt;homopolymer_nback&gt;" /> | |
| 444 <param name="preserve_qscores_less_than" type="integer" value="5" label="Bases with quality scores less than this threshold won't be recalibrated" help="-pQ,--preserve_qscores_less_than &lt;preserve_qscores_less_than&gt;"/> | |
| 445 <param name="smoothing" type="integer" value="1" label="smoothing" help="-sm,--smoothing &lt;smoothing&gt;"/> | |
| 446 <param name="max_quality_score" type="integer" value="50" label="Max quality score" help="-maxQ,--max_quality_score &lt;max_quality_score&gt;"/> | |
| 447 <param name="do_not_write_original_quals" type="boolean" checked="False" truevalue="--doNotWriteOriginalQuals" falsevalue="" label="Do Not Write Original Quality tag" help="-noOQs,--doNotWriteOriginalQuals"/> | |
| 448 </when> | |
| 449 </conditional> | |
| 450 | |
| 451 <conditional name="analysis_param_type"> | |
| 452 <param name="analysis_param_type_selector" type="select" label="Basic or Advanced Analysis options"> | |
| 453 <option value="basic" selected="True">Basic</option> | |
| 454 <option value="advanced">Advanced</option> | |
| 455 </param> | |
| 456 <when value="basic"> | |
| 457 <!-- Do nothing here --> | |
| 458 </when> | |
| 459 <when value="advanced"> | |
| 460 <param name="allow_polyploid_reduction" type="boolean" checked="False" truevalue="-polyploid" falsevalue="" label="Allow polyploid-based reduction" help="--allow_polyploid_reduction / -polyploid Allow the experimental polyploid-based reduction capabilities"/> | |
| 461 <param name="context_size" type="integer" value="10" optional="true" label="context_size" help="The number of bases to keep around mismatches (potential variation)"> | |
| 462 </param> | |
| 463 <param name="dont_compress_read_names" type="boolean" checked="False" truevalue="-nocmp_names" falsevalue="" label="Do not compress read names." help="--dont_compress_read_names / -nocmp_names By default, ReduceReads will compress read names to numbers and guarantee uniqueness and reads with similar name will still have similar compressed names. Note: If you scatter/gather there is no guarantee that read name uniqueness will be maintained -- in this case we recommend not compressing."/> | |
| 464 <param name="dont_hardclip_low_qual_tails" type="boolean" checked="False" truevalue="-noclip_tail" falsevalue="" label="Do not hard clip the low quality tails of the reads" help="--dont_hardclip_low_qual_tails / -noclip_tail This option overrides the argument of minimum tail quality"/> | |
| 465 | |
| 466 <param name="dont_simplify_reads" type="boolean" checked="False" truevalue="-nosimplify" falsevalue="" label="Do not simplify read" help="--dont_simplify_reads / -nosimplify Do not simplify read (strip away all extra information of the read -- anything other than bases, quals and read group)."/> | |
| 467 <param name="dont_use_softclipped_bases" type="boolean" checked="False" truevalue="-no_soft" falsevalue="" label="Do not use high quality soft-clipped bases" help="--dont_use_softclipped_bases / -no_soft Do not use high quality soft-clipped bases. By default, ReduceReads will hard clip away any low quality soft clipped base left by the aligner and use the high quality soft clipped bases in it's traversal algorithm to identify variant regions. The minimum quality for soft clipped bases is the same as the minimum base quality to consider (minqual)"/> | |
| 468 <param name="downsample_coverage" type="integer" value="250" optional="true" label="Downsample the coverage of a variable region" help="Downsamples the coverage of a variable region approximately (guarantees the minimum to be equal to this). A value of 0 turns downsampling off."> | |
| 469 </param> | |
| 470 <param name="hard_clip_to_interval" type="boolean" checked="False" truevalue="-clip_int" falsevalue="" label="Hard clip all incoming reads" help="--hard_clip_to_interval / -clip_int Optionally hard clip all incoming reads to the desired intervals. The hard clips will happen exactly at the interval border."/> | |
| 471 <param name="minimum_del_proportion_to_trigger_variant" type="float" value="0.05" optional="true" label="Minimum proportion of indels in a site to trigger a variant region" help="--minimum_del_proportion_to_trigger_variant / -mindel Minimum proportion of indels in a site to trigger a variant region. Anything below this will be considered consensus. "> | |
| 472 </param> | |
| 473 <param name="minimum_mapping_quality" type="integer" value="20" optional="true" label="Minimum mapping quality for consensus read" help="--minimum_mapping_quality / -minmap The minimum mapping quality to be considered for the consensus synthetic read. Reads that have mapping quality below this threshold will not be counted towards consensus, but are still counted towards variable regions."> | |
| 474 </param> | |
| 475 <param name="minimum_tail_qualities" type="integer" value="2" optional="true" label="Minimum tail quality" help="--minimum_tail_qualities / -mintail Reads have notoriously low quality bases on the tails (left and right). Consecutive bases with quality lower than this threshold will be hard clipped off before entering the reduce reads algorithm."> | |
| 476 <validator type="in_range" message="value between 0 and 127" min="0" max="127"/> | |
| 477 </param> | |
| 478 <param name="minimum_base_quality_to_consider" type="integer" value="20" optional="true" label="Minimum mapping quality for consensus read" help="--minimum_mapping_quality / -minmap The minimum mapping quality to be considered for the consensus synthetic read. Reads that have mapping quality below this threshold will not be counted towards consensus, but are still counted towards variable regions."> | |
| 479 <validator type="in_range" message="value between 0 and 127" min="0" max="127"/> | |
| 480 </param> | |
| 481 <param name="minimum_alt_proportion_to_trigger_variant" type="float" value="0.05" optional="true" label="Minimum proportion of mismatches in a site to trigger a variant region" help="--minimum_alt_proportion_to_trigger_variant / -minvar Minimum proportion of mismatches in a site to trigger a variant region. Anything below this will be considered consensus."> | |
| 482 <validator type="in_range" message="value between 0.00 and 1.00" min="0.0" max="1.0"/> | |
| 483 </param> | |
| 484 <param name="dont_hardclip_adaptor_sequences" type="boolean" checked="False" truevalue="-noclip_ad" falsevalue="" label="Do not hard clip adaptor sequences" help="--dont_hardclip_adaptor_sequences / -noclip_ad Do not hard clip adaptor sequences. Note: You don't have to turn this on for reads that are not mate paired. The program will behave correctly in those cases."/> | |
| 485 </when> | |
| 486 </conditional> | |
| 487 </inputs> | |
| 488 <outputs> | |
| 489 <data format="bam" name="output_bam" label="${tool.name} on ${on_string} (BAM)" /> | |
| 490 <data format="txt" name="output_log" label="${tool.name} on ${on_string} (log)" /> | |
| 491 </outputs> | |
| 492 <tests> | |
| 493 <test> | |
| 494 <param name="input_recal" value="gatk/gatk_count_covariates/gatk_count_covariates_out_1.csv" ftype="csv" /> | |
| 495 <param name="reference_source_selector" value="history" /> | |
| 496 <param name="ref_file" value="phiX.fasta" ftype="fasta" /> | |
| 497 <param name="input_bam" value="gatk/gatk_indel_realigner/gatk_indel_realigner_out_1.bam" ftype="bam" /> | |
| 498 <param name="gatk_param_type_selector" value="basic" /> | |
| 499 <param name="analysis_param_type_selector" value="basic" /> | |
| 500 <output name="output_bam" file="gatk/gatk_table_recalibration/gatk_table_recalibration_out_1.bam" ftype="bam" lines_diff="4" /> | |
| 501 <output name="output_log" file="gatk/gatk_table_recalibration/gatk_table_recalibration_out_1.log.contains" compare="contains" /> | |
| 502 </test> | |
| 503 </tests> | |
| 504 <help> | |
| 505 **What it does** | |
| 506 | |
| 507 ReduceReads | |
| 508 Reduces the BAM file using read based compression that keeps only essential information for variant calling | |
| 509 | |
| 510 This walker will generated reduced versions of the BAM files that still follow the BAM spec and contain all the information necessary for the GSA variant calling pipeline. Some options allow you to tune in how much compression you want to achieve. The default values have been shown to reduce a typical whole exome BAM file 100x. The higher the coverage, the bigger the savings in file size and performance of the downstream tools. | |
| 511 | |
| 512 For more information on using read based compression in the GATK, see this `tool specific page <http://www.broadinstitute.org/gatk/gatkdocs/org_broadinstitute_sting_gatk_walkers_compression_reducereads_ReduceReads.html>`_. | |
| 513 | |
| 514 To learn about best practices for variant detection using GATK, see this `overview <http://www.broadinstitute.org/gatk/guide/topic?name=best-practices>`_. | |
| 515 | |
| 516 If you encounter errors, please view the `GATK FAQ <http://www.broadinstitute.org/gatk/guide/topic?name=faqs>`_. | |
| 517 | |
| 518 ------ | |
| 519 | |
| 520 **Inputs** | |
| 521 | |
| 522 GenomeAnalysisTK: PrintReads accepts an aligned BAM and a recalibration CSV input files. | |
| 523 | |
| 524 | |
| 525 **Outputs** | |
| 526 | |
| 527 The output is in BAM format. | |
| 528 | |
| 529 | |
| 530 Go `here <http://www.broadinstitute.org/gatk/guide/topic?name=intro>`_ for details on GATK file formats. | |
| 531 | |
| 532 ------- | |
| 533 | |
| 534 **Settings**:: | |
| 535 | |
| 536 default_read_group If a read has no read group then default to the provided String. | |
| 537 default_platform If a read has no platform then default to the provided String. Valid options are illumina, 454, and solid. | |
| 538 force_read_group If provided, the read group ID of EVERY read will be forced to be the provided String. This is useful to collapse all data into a single read group. | |
| 539 force_platform If provided, the platform of EVERY read will be forced to be the provided String. Valid options are illumina, 454, and solid. | |
| 540 window_size_nqs The window size used by MinimumNQSCovariate for its calculation | |
| 541 homopolymer_nback The number of previous bases to look at in HomopolymerCovariate | |
| 542 exception_if_no_tile If provided, TileCovariate will throw an exception when no tile can be found. The default behavior is to use tile = -1 | |
| 543 solid_recal_mode How should we recalibrate solid bases in whichthe reference was inserted? Options = DO_NOTHING, SET_Q_ZERO, SET_Q_ZERO_BASE_N, or REMOVE_REF_BIAS (DO_NOTHING|SET_Q_ZERO|SET_Q_ZERO_BASE_N|REMOVE_REF_BIAS) | |
| 544 solid_nocall_strategy Defines the behavior of the recalibrator when it encounters no calls in the color space. Options = THROW_EXCEPTION, LEAVE_READ_UNRECALIBRATED, or PURGE_READ (THROW_EXCEPTION|LEAVE_READ_UNRECALIBRATED|PURGE_READ) | |
| 545 recal_file Filename for the input covariates table recalibration .csv file | |
| 546 out The output BAM file | |
| 547 bam_compression Compression level to use for writing BAM files | |
| 548 disable_bam_indexing Turn off on-the-fly creation of indices for output BAM files. | |
| 549 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 | |
| 550 preserve_qscores_less_than Bases with quality scores less than this threshold won't be recalibrated, default=5. In general it's unsafe to change qualities scores below < 5, since base callers use these values to indicate random or bad bases | |
| 551 smoothing Number of imaginary counts to add to each bin bin order to smooth out bins with few data points, default=1 | |
| 552 max_quality_score The integer value at which to cap the quality scores, default=50 | |
| 553 doNotWriteOriginalQuals If true, we will not write the original quality (OQ) tag for each read | |
| 554 | |
| 555 ------ | |
| 556 | |
| 557 **Citation** | |
| 558 | |
| 559 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>`_ | |
| 560 | |
| 561 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>`_ | |
| 562 | |
| 563 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>`_ | |
| 564 | |
| 565 </help> | |
| 566 </tool> |