From BITS wiki
Jump to: navigation, search

The framework used for the 1000 genome project, recalibrate, analyze, compare, ...

SimilarTo.png: Samtools, Bcftools, Picard, VCFtools

[ BioWare | Main_Page ]

The Genome Analysis Toolkit or GATK([1]) is a software package developed at the Broad Institute to analyse next-generation resequencing data. The toolkit offers a wide variety of tools, with a primary focus on variant discovery and genotyping as well as strong emphasis on data quality assurance. Its robust architecture, powerful processing engine and high-performance computing features make it capable of taking on projects of any size. GATK makes use of another Broad utility called Queue to perform full workflow analysis in an unsupervised manner and by applying a predefined tool sequence.

Handicon.png After registering to the Broad site, you can access a variety of reference data and tutorial at ([2]).

File:Gatk workflow.png

GATK manual page

java -jar ./GenomeAnalysisTK.jar --help
The Genome Analysis Toolkit (GATK) v2.4-9-g532efad, Compiled 2013/03/19 07:35:36
Copyright (c) 2010 The Broad Institute
For support and documentation go to
usage: java -jar GenomeAnalysisTK.jar -T <analysis_type> [-args <arg_file>] [-I <input_file>] [-rbs <read_buffer_size>] [-et
       <phone_home>] [-K <gatk_key>] [-tag <tag>] [-rf <read_filter>] [-L <intervals>] [-XL <excludeIntervals>] [-isr
       <interval_set_rule>] [-im <interval_merging>] [-ip <interval_padding>] [-R <reference_sequence>] [-ndrs]
       [--disableRandomization] [-maxRuntime <maxRuntime>] [-maxRuntimeUnits <maxRuntimeUnits>] [-dt <downsampling_type>]
       [-dfrac <downsample_to_fraction>] [-dcov <downsample_to_coverage>] [-baq <baq>] [-baqGOP <baqGapOpenPenalty>]
       [-fixMisencodedQuals] [-allowPotentiallyMisencodedQuals] [-PF <performanceLog>] [-OQ] [-BQSR <BQSR>] [-DIQ] [-EOQ]
       [-preserveQ <preserve_qscores_less_than>] [-globalQScorePrior <globalQScorePrior>] [-allowBqsrOnReducedBams] [-DBQ
       <defaultBaseQualities>] [-S <validation_strictness>] [-rpr] [-kpr] [-U <unsafe>] [-nt <num_threads>] [-nct
       <num_cpu_threads_per_data_thread>] [-mte] [-bfh <num_bam_file_handles>] [-rgbl <read_group_black_list>] [-ped
       <pedigree>] [-pedString <pedigreeString>] [-pedValidationType <pedigreeValidationType>] [-l <logging_level>] [-log
       <log_to_file>] [-h] [-version]

 -T,--analysis_type <analysis_type>                                               Type of analysis to run
 -args,--arg_file <arg_file>                                                      Reads arguments from the specified
 -I,--input_file <input_file>                                                     SAM or BAM file(s)
 -rbs,--read_buffer_size <read_buffer_size>                                       Number of reads per SAM file to buffer
                                                                                  in memory
 -et,--phone_home <phone_home>                                                    What kind of GATK run report should we
                                                                                  generate? STANDARD is the default, can
                                                                                  be NO_ET so nothing is posted to the
                                                                                  run repository. Please see
                                                                                  for details. (NO_ET|STANDARD|AWS|
 -K,--gatk_key <gatk_key>                                                         GATK Key file. Required if running
                                                                                  with -et NO_ET. Please see
                                                                                  for details.
 -tag,--tag <tag>                                                                 Arbitrary tag string to identify this
                                                                                  GATK run as part of a group of runs,
                                                                                  for later analysis
 -rf,--read_filter <read_filter>                                                  Specify filtration criteria to apply
                                                                                  to each read individually
 -L,--intervals <intervals>                                                       One or more genomic intervals over
                                                                                  which to operate. Can be explicitly
                                                                                  specified on the command line or in a
                                                                                  file (including a rod file)
 -XL,--excludeIntervals <excludeIntervals>                                        One or more genomic intervals to
                                                                                  exclude from processing. Can be
                                                                                  explicitly specified on the command
                                                                                  line or in a file (including a rod
 -isr,--interval_set_rule <interval_set_rule>                                     Indicates the set merging approach the
                                                                                  interval parser should use to combine
                                                                                  the various -L or -XL inputs (UNION|
 -im,--interval_merging <interval_merging>                                        Indicates the interval merging rule we
                                                                                  should use for abutting intervals (ALL|
 -ip,--interval_padding <interval_padding>                                        Indicates how many basepairs of
                                                                                  padding to include around each of the
                                                                                  intervals specified with the
                                                                                  -L/--intervals argument
 -R,--reference_sequence <reference_sequence>                                     Reference sequence file
 -ndrs,--nonDeterministicRandomSeed                                               Makes the GATK behave non
                                                                                  deterministically, that is, the random
                                                                                  numbers generated will be different in
                                                                                  every run
 --disableRandomization                                                           Completely eliminates randomization
                                                                                  from nondeterministic methods. To be
                                                                                  used mostly in the testing framework
                                                                                  where dynamic parallelism can result
                                                                                  in differing numbers of calls to the
 -maxRuntime,--maxRuntime <maxRuntime>                                            If provided, that GATK will stop
                                                                                  execution cleanly as soon after
                                                                                  maxRuntime has been exceeded,
                                                                                  truncating the run but not exiting
                                                                                  with a failure.  By default the value
                                                                                  is interpreted in minutes, but this
                                                                                  can be changed by maxRuntimeUnits
 -maxRuntimeUnits,--maxRuntimeUnits <maxRuntimeUnits>                             The TimeUnit for maxRuntime
 -dt,--downsampling_type <downsampling_type>                                      Type of reads downsampling to employ
                                                                                  at a given locus.  Reads will be
                                                                                  selected randomly to be removed from
                                                                                  the pile based on the method described
                                                                                  here (NONE|ALL_READS|BY_SAMPLE)
 -dfrac,--downsample_to_fraction <downsample_to_fraction>                         Fraction [0.0-1.0] of reads to
                                                                                  downsample to
 -dcov,--downsample_to_coverage <downsample_to_coverage>                          Coverage [integer] to downsample to at
                                                                                  any given locus; note that downsampled
                                                                                  reads are randomly selected from all
                                                                                  possible reads at a locus. For
                                                                                  non-locus-based traversals (eg.,
                                                                                  ReadWalkers), this sets the maximum
                                                                                  number of reads at each alignment
                                                                                  start position.
 -baq,--baq <baq>                                                                 Type of BAQ calculation to apply in
                                                                                  the engine (OFF|CALCULATE_AS_NECESSARY|
 -baqGOP,--baqGapOpenPenalty <baqGapOpenPenalty>                                  BAQ gap open penalty (Phred Scaled).
                                                                                   Default value is 40.  30 is perhaps
                                                                                  better for whole genome call sets
 -fixMisencodedQuals,--fix_misencoded_quality_scores                              Fix mis-encoded base quality scores
 -allowPotentiallyMisencodedQuals,--allow_potentially_misencoded_quality_scores   Do not fail when encountering base
                                                                                  qualities that are too high and that
                                                                                  seemingly indicate a problem with the
                                                                                  base quality encoding of the BAM file
 -PF,--performanceLog <performanceLog>                                            If provided, a GATK runtime
                                                                                  performance log will be written to
                                                                                  this file
 -OQ,--useOriginalQualities                                                       If set, use the original base quality
                                                                                  scores from the OQ tag when present
                                                                                  instead of the standard scores
 -BQSR,--BQSR <BQSR>                                                              The input covariates table file which
                                                                                  enables on-the-fly base quality score
 -DIQ,--disable_indel_quals                                                       If true, disables printing of base
                                                                                  insertion and base deletion tags (with
 -EOQ,--emit_original_quals                                                       If true, enables printing of the OQ
                                                                                  tag with the original base qualities
                                                                                  (with -BQSR)
 -preserveQ,--preserve_qscores_less_than <preserve_qscores_less_than>             Bases with quality scores less than
                                                                                  this threshold won't be recalibrated
                                                                                  (with -BQSR)
 -globalQScorePrior,--globalQScorePrior <globalQScorePrior>                       The global Qscore Bayesian prior to
                                                                                  use in the BQSR. If specified, this
                                                                                  value will be used as the prior for
                                                                                  all mismatch quality scores instead of
                                                                                  the actual reported quality score
 -allowBqsrOnReducedBams,--allow_bqsr_on_reduced_bams_despite_repeated_warnings   Do not fail when running base quality
                                                                                  score recalibration on a reduced BAM
                                                                                  file even though we highly recommend
                                                                                  against it
 -DBQ,--defaultBaseQualities <defaultBaseQualities>                               If reads are missing some or all base
                                                                                  quality scores, this value will be
                                                                                  used for all base quality scores
 -S,--validation_strictness <validation_strictness>                               How strict should we be with
                                                                                  validation (STRICT|LENIENT|SILENT)
 -rpr,--remove_program_records                                                    Should we override the Walker's
                                                                                  default and remove program records
                                                                                  from the SAM header
 -kpr,--keep_program_records                                                      Should we override the Walker's
                                                                                  default and keep program records from
                                                                                  the SAM header
 -U,--unsafe <unsafe>                                                             If set, enables unsafe operations:
                                                                                  nothing will be checked at runtime.
                                                                                   For expert users only who know what
                                                                                  they are doing.  We do not support
                                                                                  usage of this argument.
 -nt,--num_threads <num_threads>                                                  How many data threads should be
                                                                                  allocated to running this analysis.
 -nct,--num_cpu_threads_per_data_thread <num_cpu_threads_per_data_thread>         How many CPU threads should be
                                                                                  allocated per data thread to running
                                                                                  this analysis?
 -mte,--monitorThreadEfficiency                                                   Enable GATK threading efficiency
 -bfh,--num_bam_file_handles <num_bam_file_handles>                               The total number of BAM file handles
                                                                                  to keep open simultaneously
 -rgbl,--read_group_black_list <read_group_black_list>                            Filters out read groups matching
                                                                                  <TAG>:<STRING> or a .txt file
                                                                                  containing the filter strings one per
 -ped,--pedigree <pedigree>                                                       Pedigree files for samples
 -pedString,--pedigreeString <pedigreeString>                                     Pedigree string for samples
 -pedValidationType,--pedigreeValidationType <pedigreeValidationType>             How strict should we be in validating
                                                                                  the pedigree information? (STRICT|
 -l,--logging_level <logging_level>                                               Set the minimum level of logging, i.e.
                                                                                  setting INFO get's you INFO up to
                                                                                  FATAL, setting ERROR gets you ERROR
                                                                                  and FATAL level logging.
 -log,--log_to_file <log_to_file>                                                 Set the logging location
 -h,--help                                                                        Generate this help message
 -version,--version                                                               Output version information

   CheckAlignment                Validates consistency of the aligner interface by taking reads already aligned by BWA
                                 in a BAM file, stripping them of their alignment data, realigning them, and making sure
                                 one of the best resulting realignments matches the original alignment from the input
   VariantAnnotator              Annotates variant calls with context information.
   BeagleOutputToVCF             Takes files produced by Beagle imputation engine and creates a vcf with modified
   ProduceBeagleInput            Converts the input VCF into a format accepted by the Beagle imputation/analysis
   VariantsToBeagleUnphased      Produces an input file to Beagle imputation engine, listing unphased, hard-called
                                 genotypes for a single sample in input variant file.
   BaseRecalibrator              First pass of the base quality score recalibration -- Generates recalibration table
                                 based on various user-specified covariates (such as read group, reported quality score,
                                 machine cycle, and nucleotide context).
   CallableLoci                  Emits a data file containing information about callable, uncallable, poorly mapped, and
                                 other parts of the genome <p/>
   CompareCallableLoci           Test routine for new VariantContext object
   DepthOfCoverage               Toolbox for assessing sequence coverage by a wide array of metrics, partitioned by
                                 sample, read group, or library
   GCContentByInterval           Walks along reference and calculates the GC content for each interval.
   CoveredByNSamplesSites        print intervals file with all the variant sites that have "most" ( >= 90% by default)
                                 of the samples with "good" (>= 10 by default)coverage ("most" and "good" can be set in
                                 the command line).
   ErrorRatePerCycle             Computes the read error rate per position in read (in the original 5'->3' orientation
                                 that the read had coming off the machine)  Emits a GATKReport containing readgroup,
                                 cycle, mismatches, counts, qual, and error rate for each read group in the input BAMs
                                 FOR ONLY THE FIRST OF PAIR READS.
   ReadGroupProperties           Emits a GATKReport containing read group, sample, library, platform, center, sequencing
                                 data, paired end status, simple read type name (e.g.
   ReadLengthDistribution        Outputs the read lengths of all the reads in a file.
   DiffObjects                   A generic engine for comparing tree-structured objects
   GATKPaperGenotyper            A simple Bayesian genotyper, that outputs a text based call format.
   FastaAlternateReferenceMaker  Generates an alternative reference sequence over the specified interval.
   FastaReferenceMaker           Renders a new reference in FASTA format consisting of only those loci provided in the
                                 input data set.
   FastaStats                    Calculates basic statistics about the reference sequence itself
   VariantFiltration             Filters variant calls using a number of user-selectable, parameterizable criteria.
   UnifiedGenotyper              A variant caller which unifies the approaches of several disparate callers -- Works for
                                 single-sample and multi-sample data.
   HaplotypeCaller               Call SNPs and indels simultaneously via local de-novo assembly of haplotypes in an
                                 active region.
   HaplotypeResolver             Haplotype-based resolution of variants in 2 different eval files.
   IndelRealigner                Performs local realignment of reads based on misalignments due to the presence of
   LeftAlignIndels               Left-aligns indels from reads in a bam file.
   RealignerTargetCreator        Emits intervals for the Local Indel Realigner to target for realignment.
   PhaseByTransmission           Computes the most likely genotype combination and phases trios and parent/child pairs
   ReadBackedPhasing             Walks along all variant ROD loci, caching a user-defined window of VariantContext
                                 sites, and then finishes phasing them when they go out of range (using upstream and
                                 downstream reads).
   CheckPileup                   At every locus in the input set, compares the pileup data (reference base, aligned base
                                 from each overlapping read, and quality score) to the reference pileup data generated
                                 by samtools.
   CountBases                    Walks over the input data set, calculating the number of bases seen for diagnostic
   CountIntervals                Counts the number of contiguous regions the walker traverses over.
   CountLoci                     Walks over the input data set, calculating the total number of covered loci for
                                 diagnostic purposes.
   CountMales                    Walks over the input data set, calculating the number of reads seen from male samples
                                 for diagnostic purposes.
   CountReadEvents               Walks over the input data set, counting the number of read events (from the CIGAR
   CountReads                    Walks over the input data set, calculating the number of reads seen for diagnostic
   CountRODs                     Prints out counts of the number of reference ordered data objects encountered.
   CountRODsByRef                Prints out counts of the number of reference ordered data objects encountered.
   CountTerminusEvent            Walks over the input data set, counting the number of reads ending in
                                 insertions/deletions or soft-clips
   FlagStat                      A reimplementation of the 'samtools flagstat' subcommand in the GATK.
   Pileup                        Prints the alignment in something similar to the samtools pileup format.
   PrintRODs                     Prints out all of the RODs in the input data set.
   QCRef                         Quality control for the reference fasta
   ReadClippingStats             Walks over the input reads, printing out statistics about the read length, number of
                                 clipping events, and length of the clipping to the output stream.
   ClipReads                     This tool provides simple, powerful read clipping capabilities to remove low quality
                                 strings of bases, sections of reads, and reads containing user-provided sequences.
   PrintReads                    Renders, in SAM/BAM format, all reads from the input data set in the order in which
                                 they appear in the input file.
   SplitSamFile                  Divides the input data set into separate BAM files, one for each sample in the input
                                 data set.
   CompareBAM                    Given two BAMs with different read groups, it compares them based on ReduceReads
   ReduceReads                   Reduces the BAM file using read based compression that keeps only essential information
                                 for variant calling
   BaseCoverageDistribution      Simple walker to plot the coverage distribution per base.
   DiagnoseTargets               Analyzes coverage distribution and validates read mates for a given interval and
   GenotypeAndValidate           Genotypes a dataset and validates the calls of another dataset using the Unified
   ValidationAmplicons           Creates FASTA sequences for use in Seqenom or PCR utilities for site amplification and
                                 subsequent validation
   ValidationSiteSelector        Randomly selects VCF records according to specified options.
   VariantEval                   General-purpose tool for variant evaluation (% in dbSNP, genotype concordance, Ti/Tv
                                 ratios, and a lot more)
   ApplyRecalibration            Applies cuts to the input vcf file (by adding filter lines) to achieve the desired
                                 novel truth sensitivity levels which were specified during VariantRecalibration
   VariantRecalibrator           Create a Gaussian mixture model by looking at the annotations values over a high
                                 quality subset of the input call set and then evaluate all input variants.
   CombineVariants               Combines VCF records from different sources.
   FilterLiftedVariants          Filters a lifted-over VCF file for ref bases that have been changed.
   GenotypeConcordance           A simple walker for performing genotype concordance calculations between two callsets.
   LeftAlignVariants             Left-aligns indels from a variants file.
   LiftoverVariants              Lifts a VCF file over from one build to another.
   RandomlySplitVariants         Takes a VCF file, randomly splits variants into two different sets, and outputs 2 new
                                 VCFs with the results.
   RegenotypeVariants            Regenotypes the variants from a VCF.
   SelectHeaders                 Selects headers from a VCF source.
   SelectVariants                Selects variants from a VCF source.
   ValidateVariants              Validates a VCF file with an extra strict set of criteria.
   VariantsToBinaryPed           Converts a VCF file to a binary plink Ped file (.bed/.bim/.fam)
   VariantsToTable               Emits specific fields from a VCF file to a tab-deliminated table
   VariantsToVCF                 Converts variants from other file formats to VCF format.
   VariantValidationAssessor     Annotates a validation (from Sequenom for example) VCF with QC metrics (HW-equilibrium,
                                 % failed probes)


[ BioWare | Main_Page ]