org.broadinstitute.hellbender.engine

Class DuplicateSetWalker

java.lang.Object

org.broadinstitute.hellbender.cmdline.CommandLineProgram

org.broadinstitute.hellbender.engine.GATKTool

org.broadinstitute.hellbender.engine.WalkerBase

org.broadinstitute.hellbender.engine.ReadWalker

org.broadinstitute.hellbender.engine.DuplicateSetWalker

All Implemented Interfaces:

org.broadinstitute.barclay.argparser.CommandLinePluginProvider

Direct Known Subclasses:

DownsampleByDuplicateSet

public abstract class DuplicateSetWalker
extends ReadWalker

A walker that processes duplicate reads that share the same Unique molecule Identifier (UMI) as a single unit. This tool assumes that the input bam has been sorted by UMI (the SAMTag.MI tag to be specific) with FGBio GroupReadsByUmi: http://fulcrumgenomics.github.io/fgbio/tools/latest/GroupReadsByUmi.html

Field Summary

Fields

Modifier and Type	Field and Description
protected ReadsWithSameUMI	currentReadsWithSameUMI
static java.lang.String	MIN_REQUIRED_READS_NAME
static java.lang.String	MIN_REQUIRED_READS_PER_STRAND_NAME

Fields inherited from class org.broadinstitute.hellbender.engine.ReadWalker

FEATURE_CACHE_LOOKAHEAD

Fields inherited from class org.broadinstitute.hellbender.engine.GATKTool

addOutputSAMProgramRecord, addOutputVCFCommandLine, cloudIndexPrefetchBuffer, cloudPrefetchBuffer, createOutputBamIndex, createOutputBamMD5, createOutputVariantIndex, createOutputVariantMD5, disableBamIndexCaching, features, intervalArgumentCollection, lenientVCFProcessing, outputSitesOnlyVCFs, progressMeter, readArguments, referenceArguments, SECONDS_BETWEEN_PROGRESS_UPDATES_NAME, seqValidationArguments

Fields inherited from class org.broadinstitute.hellbender.cmdline.CommandLineProgram

GATK_CONFIG_FILE, logger, NIO_MAX_REOPENS, NIO_PROJECT_FOR_REQUESTER_PAYS, QUIET, specialArgumentsCollection, tmpDir, useJdkDeflater, useJdkInflater, VERBOSITY

Constructor Summary

Constructors

Constructor and Description
DuplicateSetWalker()

Method Summary

Modifier and Type	Method and Description
void	apply(GATKRead read, ReferenceContext referenceContext, FeatureContext featureContext) FGBio GroupByUMI returns reads sorted by molecule ID: For example, the input bam may look like read1: ...
abstract void	apply(ReadsWithSameUMI readsWithSameUMI, ReferenceContext referenceContext, FeatureContext featureContext) A subclass must specify how to process the duplicate sets by overriding this method.
java.util.List<ReadFilter>	getDefaultReadFilters() Returns the default list of CommandLineReadFilters that are used for this tool.
protected boolean	rejectSet(ReadsWithSameUMI readsWithSameUMI) Returns true for duplicate sets that does not meet required criteria for further processing.
void	traverse() A complete traversal from start to finish.

Methods inherited from class org.broadinstitute.hellbender.engine.ReadWalker

getProgressMeterRecordLabel, onShutdown, onStartup, requiresReads, resetReadsDataSource

Methods inherited from class org.broadinstitute.hellbender.engine.WalkerBase

directlyAccessEngineFeatureManager, directlyAccessEngineReadsDataSource, directlyAccessEngineReferenceDataSource

Methods inherited from class org.broadinstitute.hellbender.engine.GATKTool

addFeatureInputsAfterInitialization, bamIndexCachingShouldBeEnabled, closeTool, createSAMWriter, createVCFWriter, createVCFWriter, createVCFWriter, disableProgressMeter, doWork, getBestAvailableSequenceDictionary, getDefaultCloudIndexPrefetchBufferSize, getDefaultCloudPrefetchBufferSize, getDefaultToolVCFHeaderLines, getDefaultVariantAnnotationGroups, getDefaultVariantAnnotations, getGenomicsDBOptions, getHeaderForFeatures, getHeaderForReads, getHeaderForSAMWriter, getMasterSequenceDictionary, getPluginDescriptors, getReferenceDictionary, getSequenceDictionaryValidationArgumentCollection, getToolName, getTransformedReadStream, getTraversalIntervals, hasFeatures, hasReads, hasReference, hasUserSuppliedIntervals, initializeProgressMeter, makePostReadFilterTransformer, makePreReadFilterTransformer, makeReadFilter, makeSamReaderFactory, makeVariantAnnotations, onTraversalStart, onTraversalSuccess, requiresFeatures, requiresIntervals, requiresReference, transformTraversalIntervals, useVariantAnnotations

Methods inherited from class org.broadinstitute.hellbender.cmdline.CommandLineProgram

customCommandLineValidation, getCommandLine, getCommandLineParser, getDefaultHeaders, getMetricsFile, getSupportInformation, getToolkitName, getToolkitShortName, getToolStatusWarning, getUsage, getVersion, instanceMain, instanceMainPostParseArgs, isBetaFeature, isExperimentalFeature, parseArgs, printLibraryVersions, printSettings, printStartupMessage, runTool, setDefaultHeaders, warnOnToolStatus

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

MIN_REQUIRED_READS_NAME

public static final java.lang.String MIN_REQUIRED_READS_NAME

See Also:

Constant Field Values

MIN_REQUIRED_READS_PER_STRAND_NAME

public static final java.lang.String MIN_REQUIRED_READS_PER_STRAND_NAME

See Also:

Constant Field Values

currentReadsWithSameUMI

protected ReadsWithSameUMI currentReadsWithSameUMI

Constructor Detail

DuplicateSetWalker

public DuplicateSetWalker()

Method Detail

traverse

public final void traverse()

Description copied from class: ReadWalker

A complete traversal from start to finish. Tool authors who wish to "roll their own" traversal from scratch can extend this class directly and implement this method. Walker authors should instead extend a Walker class and implement the Walker-appropriate apply() method, since the Walker base classes implement the various kinds of traversals for you. Implementation of read-based traversal. The default implementation creates filters using GATKTool.makeReadFilter() and transformers using GATKTool.makePreReadFilterTransformer() GATKTool.makePostReadFilterTransformer() and then iterates over all reads, applies the pre-filter transformer, the filter, then the post-filter transformer and hands the resulting reads to the ReadWalker.apply(org.broadinstitute.hellbender.utils.read.GATKRead, org.broadinstitute.hellbender.engine.ReferenceContext, org.broadinstitute.hellbender.engine.FeatureContext) function of the walker (along with additional contextual information, if present, such as reference bases). NOTE: You should only override ReadWalker.traverse() if you are writing a new walker base class in the engine package that extends this class. It is not meant to be overridden by tools outside of the engine package.

Overrides:

traverse in class ReadWalker

apply

public final void apply(GATKRead read,
                        ReferenceContext referenceContext,
                        FeatureContext featureContext)

FGBio GroupByUMI returns reads sorted by molecule ID: For example, the input bam may look like read1: ... MI:Z:0/A ... read2: ... MI:Z:0/A ... read3: ... MI:Z:0/B ... read4: ... MI:Z:0/B ... read5: ... MI:Z:1/A ... read6: ... MI:Z:1/B ... read7: ... MI:Z:1/B ... Thus it's sufficient to go through the reads in order and collect them in a list until we encounter the next molecule ID, at which point we pass the list to the apply method, process the set based on the child class's implementation of the method, and clear the currentDuplicateSet variable and start collecting reads again. Notice there are two apply() methods in this class: This apply() inherited from ReadWalker is marked final to discourage subclassing. A subclass must override the other apply() method that takes in the DuplicateSet.

Specified by:

apply in class ReadWalker

Parameters:

read - current read

referenceContext - Reference bases spanning the current read. Will be an empty, but non-null, context object if there is no backing source of reference data (in which case all queries on it will return an empty array/iterator). Can request extra bases of context around the current read's interval by invoking ReferenceContext.setWindow(int, int) on this object before calling ReferenceContext.getBases()

featureContext - Features spanning the current read. Will be an empty, but non-null, context object if there is no backing source of Feature data (in which case all queries on it will return an empty List).

apply

public abstract void apply(ReadsWithSameUMI readsWithSameUMI,
                           ReferenceContext referenceContext,
                           FeatureContext featureContext)

A subclass must specify how to process the duplicate sets by overriding this method.

Parameters:

readsWithSameUMI - A set of reads with the matching UMIs with the same fragment start and end

referenceContext - A reference context object over the intervals determined by the duplicate set.

featureContext - Entries from a secondary feature file (e.g. vcf) if provided

rejectSet

protected boolean rejectSet(ReadsWithSameUMI readsWithSameUMI)

Returns true for duplicate sets that does not meet required criteria for further processing. We encourage the user to override this method to meet their needs.

getDefaultReadFilters

public java.util.List<ReadFilter> getDefaultReadFilters()

Description copied from class: ReadWalker

Returns the default list of CommandLineReadFilters that are used for this tool. The filters returned by this method are subject to selective enabling/disabling by the user via the command line. The default implementation uses the WellformedReadFilter filter with all default options. Subclasses can override to provide alternative filters. Note: this method is called before command line parsing begins, and thus before a SAMFileHeader is available through {link #getHeaderForReads}.

Overrides:

getDefaultReadFilters in class ReadWalker

Returns:

List of individual filters to be applied for this tool.