Mercurial > repos > iuc > gatk2
diff haplotype_caller.xml @ 2:8bcc13094767 draft
Uploaded
| author | iuc |
|---|---|
| date | Sat, 18 Jan 2014 07:21:33 -0500 |
| parents | 340633249b3d |
| children | f244b8209eb8 |
line wrap: on
line diff
--- a/haplotype_caller.xml Mon Dec 02 10:36:02 2013 -0500 +++ b/haplotype_caller.xml Sat Jan 18 07:21:33 2014 -0500 @@ -7,10 +7,12 @@ <command interpreter="python"> gatk2_wrapper.py --stdout "${output_log}" - -d "-I" "${reference_source.input_bam}" "${reference_source.input_bam.ext}" "gatk_input" - #if str( $reference_source.input_bam.metadata.bam_index ) != "None": - -d "" "${reference_source.input_bam.metadata.bam_index}" "bam_index" "gatk_input" ##hardcode galaxy ext type as bam_index - #end if + #for $i, $input_bam in enumerate( $reference_source.input_bams ): + -d "-I" "${input_bam.input_bam}" "${input_bam.input_bam.ext}" "gatk_input_${i}" + #if str( $input_bam.input_bam.metadata.bam_index ) != "None": + -d "" "${input_bam.input_bam.metadata.bam_index}" "bam_index" "gatk_input_${i}" ##hardcode galaxy ext type as bam_index + #end if + #end for -p ' @JAR_PATH@ -T "HaplotypeCaller" @@ -18,7 +20,7 @@ \$GATK2_SITE_OPTIONS - @THREADS@ + --num_cpu_threads_per_data_thread \${GALAXY_SLOTS:-4} #if $reference_source.reference_source_selector != "history": -R "${reference_source.ref_file.fields.path}" @@ -27,14 +29,12 @@ --BQSR "${input_recal}" #end if ' + @DBSNP_OPTIONS@ #include source=$standard_gatk_options# ##start analysis specific options #if $analysis_param_type.analysis_param_type_selector == "advanced": -p ' - #if $analysis_param_type.p_nonref_model.__str__ != "None" and len($analysis_param_type.p_nonref_model.__str__) > 0: - --p_nonref_model $analysis_param_type.p_nonref_model - #end if #if $analysis_param_type.heterozygosity.__str__.strip() != '': --heterozygosity $analysis_param_type.heterozygosity #end if @@ -42,8 +42,8 @@ #if str( $analysis_param_type.genotyping_mode_type.genotyping_mode ) == 'GENOTYPE_GIVEN_ALLELES': --alleles "${analysis_param_type.genotyping_mode_type.input_alleles_rod}" #end if - #if $analysis_param_type.output_mode.__str__ != "None" and len($analysis_param_type.output_mode.__str__) > 0: - --output_mode $analysis_param_type.output_mode + #if not $analysis_param_type.emitRefConfidence is None: + --emitRefConfidence $analysis_param_type.emitRefConfidence #end if ## files @@ -53,9 +53,6 @@ #if str($analysis_param_type.comp) != 'None': --comp "$analysis_param_type.comp" #end if - #if str($analysis_param_type.dbsnp) != 'None': - --dbsnp "$analysis_param_type.dbsnp" - #end if ## #if str( $analysis_param_type.annotation ) != "None": #for $annotation in str( $analysis_param_type.annotation.fields.gatk_value ).split( ','): @@ -80,9 +77,6 @@ #if $analysis_param_type.contamination_fraction_to_filter.__str__.strip() != '': --contamination_fraction_to_filter $analysis_param_type.contamination_fraction_to_filter #end if - #if $analysis_param_type.downsampleRegion.__str__.strip() != '': - --downsampleRegion $analysis_param_type.downsampleRegion - #end if #if $analysis_param_type.minPruning.__str__.strip() != '': --minPruning $analysis_param_type.minPruning #end if @@ -99,9 +93,7 @@ --max_alternate_alleles $analysis_param_type.max_alternate_alleles #end if ## mode selections - #if $analysis_param_type.genotyping_mode.__str__ != "None" and len($analysis_param_type.genotyping_mode.__str__) > 0: - --genotyping_mode $analysis_param_type.genotyping_mode - #end if + #if $analysis_param_type.pair_hmm_implementation.__str__ != "None" and len($analysis_param_type.pair_hmm_implementation.__str__) > 0: --pair_hmm_implementation $analysis_param_type.pair_hmm_implementation #end if @@ -121,35 +113,31 @@ #end if </command> <inputs> - <param name="input_recal" type="data" format="gatk_report" optional="true" label="Covariates table recalibration file" help="-BQSR,--BQSR &lt;recal_file&gt;" > - <help>The input covariates table file which enables on-the-fly base quality score recalibration. - Enables on-the-fly recalibrate of base qualities. The covariates tables are produced by the BaseQualityScoreRecalibrator tool. - Please be aware that one should only run recalibration with the covariates file created on the same input bam(s). - </help> - </param> + <param name="input_recal" type="data" format="gatk_report" optional="true" label="Covariates table recalibration file" help="The input covariates table file which enables on-the-fly base quality score recalibration. Enables on-the-fly recalibrate of base qualities. The covariates tables are produced by the BaseQualityScoreRecalibrator tool. Please be aware that one should only run recalibration with the covariates file created on the same input bam(s) (-BQSR,--BQSR &lt;recal_file&gt;)" /> <conditional name="reference_source"> <expand macro="reference_source_selector_param" /> <when value="cached"> - <param name="input_bam" type="data" format="bam" label="BAM file" help="-I,--input_file &lt;input_file&gt;"> - <validator type="unspecified_build" /> - <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 --> - </param> + <repeat name="input_bams" title="BAM file" min="1" help="-I,--input_file &lt;input_file&gt;"> + <param name="input_bam" type="data" format="bam" label="BAM file"> + <validator type="unspecified_build" /> + <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 --> + </param> + </repeat> <param name="ref_file" type="select" label="Using reference genome" help="-R,--reference_sequence &lt;reference_sequence&gt;" > <options from_data_table="gatk2_picard_indexes"> - <filter type="data_meta" key="dbkey" ref="input_bam" column="dbkey"/> + <!-- <filter type="data_meta" key="dbkey" ref="input_bam" column="dbkey"/> does not yet work in a repeat...--> </options> <validator type="no_options" message="A built-in reference genome is not available for the build associated with the selected input file"/> </param> </when> <when value="history"> - <param name="input_bam" type="data" format="bam" label="BAM file" help="-I,--input_file &lt;input_file&gt;" /> - <param name="ref_file" type="data" format="fasta" label="Using reference file" help="-R,--reference_sequence &lt;reference_sequence&gt;"> - <options> - <filter type="data_meta" key="dbkey" ref="input_bam" /> - </options> - </param> + <repeat name="input_bams" title="BAM file" min="1" help="-I,--input_file &lt;input_file&gt;"> + <param name="input_bam" type="data" format="bam" label="BAM file" /> + </repeat> + <param name="ref_file" type="data" format="fasta" label="Using reference file" help="-R,--reference_sequence &lt;reference_sequence&gt;" /> </when> </conditional> + <expand macro="dbsnp_param" /> <expand macro="gatk_param_type_conditional" /> @@ -211,9 +199,7 @@ <param name="contamination_fraction_to_filter" type="float" value="0.05" optional="true" label="contamination_fraction_to_filter" help="--contamination_fraction_to_filter / -contamination Fraction of contamination in sequencing data (for all samples) to aggressively remove"> <validator type="in_range" message="value between 0.00 and 1.00" min="0" max="1"/> </param> - <param name="dbsnp" type="data" format="vcf" optional="true" label="dbsnp" help="--dbsnp / -D dbSNP file"/> <param name="debug" type="boolean" checked="False" truevalue="-debug" falsevalue="" label="debug" help="--debug / -debug If specified, print out very verbose debug information about each triggering active region"/> - <param name="downsampleRegion" type="integer" value="1000" optional="true" label="downsampleRegion" help="--downsampleRegion / -dr coverage, per-sample, to downsample each active region to"/> <conditional name="genotyping_mode_type"> <param name="genotyping_mode" type="select" label="How to determine the alternate allele to use for genotyping" help="-gt_mode,--genotyping_mode &lt;genotyping_mode&gt;"> @@ -234,10 +220,11 @@ <param name="minPruning" type="integer" value="1" optional="true" label="minPruning" help="--minPruning / -minPruning The minimum allowed pruning factor in assembly graph. Paths with >= X supporting kmers are pruned from the graph"> <validator type="in_range" message="value between 0 and 127" min="0" max="127"/> </param> - <param name="output_mode" type="select" optional="true" label="output_mode" help="--output_mode / -out_mode Specifies which type of calls we should output"> - <option value="EMIT_VARIANTS_ONLY" selected="True">EMIT_VARIANTS_ONLY</option> - <option value="EMIT_ALL_CONFIDENT_SITES">EMIT_ALL_CONFIDENT_SITES</option> - <option value="EMIT_ALL_SITES">EMIT_ALL_SITES</option> + <!-- http://www.broadinstitute.org/gatk/guide/article?id=2940 --> + <param name="emitRefConfidence" type="select" optional="true" label="Output confidence estimates" help="Emitting a per-bp or summarized confidence estimate for a site being strictly homozygous-reference (--emitRefConfidence)"> + <option value="NONE" selected="True">don't emit anything</option> + <option value="BP_RESOLUTION">BP_RESOLUTION (emit detailed information for each BP)</option> + <option value="GVCF">GVCF (emit a block summarized version of the BP_RESOLUTION data)</option> </param> <param name="pair_hmm_implementation" type="select" optional="true" label="pair_hmm_implementation" help="--pair_hmm_implementation / -pairHMM The PairHMM implementation to use for genotype likelihood calculations"> <option value="EXACT">EXACT</option> @@ -252,13 +239,6 @@ <param name="gcpHMM" type="integer" value="10" optional="true" label="gcpHMM" help="--gcpHMM / -gcpHMM Flat gap continuation penalty for use in the Pair HMM"/> <param name="genotypeFullActiveRegion" type="boolean" checked="False" truevalue="-genotypeFullActiveRegion" falsevalue="" label="genotypeFullActiveRegion" help="--genotypeFullActiveRegion / -genotypeFullActiveRegion If specified, alternate alleles are considered to be the full active region for the purposes of genotyping"/> <param name="max_alternate_alleles" type="integer" value="6" optional="true" label="max_alternate_alleles" help="--max_alternate_alleles / -maxAltAlleles Maximum number of alternate alleles to genotype"/> - <param name="p_nonref_model" type="select" optional="true" label="p_nonref_model" help="--p_nonref_model / -pnrm Non-reference probability calculation model to employ"> - <option value="EXACT_INDEPENDENT" selected="True">EXACT_INDEPENDENT experimental implementation - for testing only</option> - <option value="EXACT_REFERENCE">EXACT_REFERENCE reference implementation of multi-allelic EXACT model. Extremely slow for many alternate alleles</option> - <option value="EXACT_ORIGINAL">EXACT_ORIGINAL original biallelic exact model, for testing only</option> - <option value="EXACT_GENERAL_PLOIDY">implementation that supports any sample ploidy</option> - </param> - </when> </conditional> </inputs> @@ -323,14 +303,12 @@ contamination Fraction of contamination in sequencing data (for all samples) to aggressively remove dbsnp dbSNP file debug If specified, print out very verbose debug information about each triggering active region - downsampleRegion coverage, per-sample, to downsample each active region to excludeAnnotation One or more specific annotations to exclude genotyping_mode Specifies how to determine the alternate alleles to use for genotyping graphOutput File to which debug assembly graph information should be written group One or more classes/groups of annotations to apply to variant calls heterozygosity Heterozygosity value used to compute prior likelihoods for any locus minPruning The minimum allowed pruning factor in assembly graph. Paths with less than or equal supporting kmers are pruned from the graph - output_mode Specifies which type of calls we should output pair_hmm_implementation The PairHMM implementation to use for genotype likelihood calculations stand_call_conf The minimum phred-scaled confidence threshold at which variants should be called stand_emit_conf The minimum phred-scaled confidence threshold at which variants should be emitted (and filtered with LowQual if less than the calling threshold) @@ -339,9 +317,6 @@ gcpHMM Flat gap continuation penalty for use in the Pair HMM genotypeFullActiveRegion If specified, alternate alleles are considered to be the full active region for the purposes of genotyping max_alternate_alleles Maximum number of alternate alleles to genotype - p_nonref_model Non-reference probability calculation model to employ - ------- @CITATION_SECTION@ </help>