1) Better documentation on the meta data file for VariantsToBinaryPed with examples of each file type
2) MannWhitneyU can now take an argument on creation to turn off dithering. This pertains to JIRA-GSA-571 but does not fix it,
as it isn't hooked up to the command line. Next step is to add an argument to the command line where it's accessible to the
annotation classes (e.g. from either UG or the VariantAnnotator).
3) Added some dumb python scripts to deal with Plink files, and a script to convert plink binaries to VCF to help sanity check. Basically if you want to do an analysis on genotype data stored in plink binary format, your choices are:
1) Add a new module to Plink [difficulty rating: Impossible -- code obfuscation]
2) Steal plink parsing code from software (Plink/PlinkSeq/GCTA/Emacks/etc) that readds the files [difficulty rating: Oppressive -- code not modularized at all)
3) Write your own dumb stuff [difficutly rating: Annoying]
What's been added is the result of 3. It's a library so nobody else has to do this, so long as they're comfortable with python.
-- Renamed TraversalErrorManager to the more general MultiThreadedErrorTracker
-- ErrorTracker is now used throughout the NanoScheduler. In order to properly handle errors, the work previously done by main thread (submit jobs, block on reduce) is now handled in a separate thread. The main thread simply wakes up peroidically and checks whether the reduce result is available or if an error has occurred, and handles each appropriately.
-- EngineFeaturesIntegrationTest checks that -nt and -nct properly throw errors in Walkers
-- Added NanoSchedulerUnitTest for input errors
-- ThreadEfficiencyMonitoring is now disabled by default, and can be enabled with a GATK command line option. This is because the monitoring doesn't differentiate between threads that are supposed to do work, and those that are supposed to wait, and therefore gives misleading results.
-- Build.xml no longer copies the unittest results verbosely
-- Refactored error handling from HMS into utils.TraversalErrorManager, which is now used by HMS and will be usable by NanoScheduler
-- Generalized EngineFeaturesIntegrationTest to test map / reduce error throwing for nt 1, nt 2 and nct 2 (disabled)
-- Added unit tests for failing input iterator in NanoScheduler (fails)
-- Made ErrorThrowing NanoScheduable
-- V3 + V4 algorithm for NanoScheduler. The newer version uses 1 dedicated input thread and n - 1 map/reduce threads. These MapReduceJobs perform map and a greedy reduce. The main thread's only job is to shuttle inputs from the input producer thread, enqueueing MapReduce jobs for each one. We manage the number of map jobs now via a Semaphore instead of a BlockingQueue of fixed size.
-- This new algorithm should consume N00% CPU power for -nct N value.
-- Also a cleaner implementation in general
-- Vastly expanded unit tests
-- Deleted FutureValue and ReduceThread
-- Turns out this was consuming 30% of the UG runtime, and causing problems elsewhere.
-- Removed addMissingSamples from VariantcontextUtils, and calls to it
-- Updated VCF / BCF writers to automatically write out a diploid no call for missing samples
-- Added unit tests for this behavior in VariantContextWritersUnitTest
1) SelectVariants could throw a ReviewedStingException (one of the nasty "Bug:") ones if the user requested a sample that wasn't present in the VCF. The walker now
checks for this in the initialize() phase, and throws a more informative error if the situation is detected. If the user simply wants to subset the VCF to
all the samples requested that are actually present in the VCF, the --ALLOW_NONOVERLAPPING_COMMAND_LINE_SAMPLES flag changes this UserException to a Warning,
and does the appropriate subsetting. Added integration tests for this.
2) GenotypeLikelihoods has an unsafe method getLog10GQ(GenotypeType), which is completely broken for multi-allelic sites. I marked that method
as deprecated, and added methods that use the context of the allele ordering (either directly specified or as a VC) to retrieve the appropriate GQ, and
added a unit test to cover this case. VariantsToBinaryPed needs to dynamically calculate the GQ field sometimes (because I have some VCFs with PLs but no GQ).
-- Now prints out a single combined NanoScheduler runtime profile report across all nano schedulers in use. So now if you run with -nt 4 you'll get one combined NanoScheduler profiler across all 4 instances of the NanoScheduler within TraverseXNano.
-- Previously these core progress metering functions were all in TraversalEngine, and available to subclasses like TraverseLoci via inheritance. The problem here is that the upcoming data threads x cpu threads parallelism requires one master copy of the progress metering shared among all traversals, but multiple instantiations of traverse engines themselves.
-- Because the progress metering code has horrible anyway, I've refactored and vastly cleaned up and simplified all of these capabilities into TraversalProgressMeter class. I've simplified down the classes it uses to work (STILL SOME TODOs in there) so that it doesn't reach into the core GATK engine all the time. It should be possible to write some nice tests for it now. By making it its own class, it can protect itself from multi-threaded access with a single synchronized printProgress function instead of carrying around multiple lock objects as before
-- Cleaned up the start up of the progress meter. It's now handled when the meter is created, so each micro scheduler doesn't have to deal with proper initialization timing any longer
-- Simplified and made clear the interface for shutting down the traversal engines. There's no a shutdown method in TraversalEngine that's called once by the MicroScheduler when the entire traversing in over. Nano traversals now properly shut down (was subtle bug I undercovered here). The printing of on traversal done metering is now handled by MicroScheduler
-- The MicroScheduler holds the single master copy of the progress meter, and doles it out to the TraversalEngines (currently 1 but in future commit there will be N).
-- Added a nice function to GenomeAnalysisEngine that returns the regions we will be processing, either the intervals requested or the whole genome. Useful for progress meter but also probably for other infrastructure as well
-- Remove a lot of the sh*ting Bean interface getting and setting in MicroScheduler that's no longer useful. The generic bean is just a shell interface with nothing in it.
-- By removing a lot of these bean accessors and setters many things are now final that used to be dynamic.
-- I've rewritten the entire NS framework to use a producer / consumer model for input -> map and from map -> reduce. This is allowing us to scale reasonably efficiently up to 4 threads (see figure). Future work on the nano scheduler will be itemized in a separate JIRA entry.
-- Restructured the NS code for clarity. Docs everywhere.
-- This is considered version 1.0
-Off by default; engine fork isolates new code paths from old code paths,
so no integration tests change yet
-Experimental implementation is currently BROKEN due to a serious issue
involving file spans. No one can/should use the experimental features
until I've patched this issue.
-There are temporarily two independent versions of LocusIteratorByState.
Anyone changing one version should port the change to the other (if possible),
and anyone adding unit tests for one version should add the same unit tests
for the other (again, if possible). This situation will hopefully be extremely
temporary, and last only until the experimental implementation is proven.
-- Separate updating cumulative traversal metrics from printing progress. There's now an updateCumulativeMetrics function and a printProgress() that only takes a current position
-- printProgress now soles relies on the time since the last progress to decide if it will print or not. No longer uses the number of cycles, since this isn't reliable in the case of nano scheduling
-- GenomeAnalysisEngine now maintains a pointer to the master cumulative metrics. getCumulativeMetrics never returns null, which was handled in some parts of the code but not others.
-- Update all of the traversals to use the new updateCumulativeMetrics, printProgress model
-- Added progress callback to nano scheduler. Every bufferSize elements this callback is invoked, allowing us to smoothly update the progress meter in the NanoScheduler
-- Rename MapFunction to NanoSchedulerMap and the same for reduce.
- VariantAnnotatorEngine changed to call genotype annotations even if pilups and allele -> likelihood mappings are not present. Current genotype annotations altered to check for null pilupes and null mappings.
-- Confirmed that reads spanning off the end of the chromosome don't cause an exception by adding integration test for a single read that starts 7 bases from the end of chromosome 1 and spans 90 bases or so off. Added pileup integration test to ensure this behavior continues to work
-- Yes, GenomeLoc.compareTo was broken. The compareTo function only considered the contig and start position, but not the stop, when comparing genome locs.
-- Updated GenomeLoc.compareTo function to account for stop. Updated GATK code where necessary to fix resulting problems that depended on this.
-- Added unit tests to ensure that hashcode, equals, and compareTo are all correct for GenomeLocs
-- Stateless objects are required for nano-scheduling. This means you can take the RefMetaDataTracker provided by ReadBasedReferenceOrderedView, store it way, get another from the same view, and the original one behaves the same.
-- Previous behavior was unnecessary and causes all sorts of problems with RODs for reads. The old implementation simply failed in this case. The new code handles this correctly by forcing shards to have all of their data on a single contig.
-- Added a PrintReads integration test to ensure this behavior is correct
-- Adding test BAMs that have < 200 reads and span across contig boundaries
-- Deleted ReadMetaDataTracker
-- Added function to ReadShard to give us the span from the left most position of the reads in the shard to the right most, which is needed for the new view
-- ReadMetaDataTracker is dead! Long live the RefMetaDataTracker. Read walkers will soon just take RefMetaDataTracker objects. In this commit they take a class that trivially extends them
-- Rewrote ReadBasedReferenceOrderedView to produce RefMetaDataTrackers not the old class.
-- This new implementation produces thread-safe objects (i.e., holds no points to shared state). Suitable for use (to be tested) with nano scheduling
-- Simplified interfaces to use the simplest data structures (PeekableIterator) not the LocusAwareSeekableIterator, since I both hate those classes and this is on the long term trajectory to remove those from the GATK entirely.
-- Massively expanded DataProvider unit tests for ReadBasedReferenceOrderedView
-- Note that the old implementation of offset -> ROD in ReadRefMetaDataTracker was broken for any read not completely matching the reference. Rather than provide broken code the ReadMetaDataTracker only provides a "bag of RODs" interface. If you want to work with the relationship between the read and the RODs in your tool you need to manage the CIGAR element itself.
-- This commit breaks the new read walker BQSR, but Ryan knows this is coming
-- Subsequent commit will be retiring / fixing ValidateRODForReads
-- Groups inputs for each thread so that we don't have one thread execution per map() call
-- Added shutdown function
-- Documentation everywhere
-- Code cleanup
-- Extensive unittests
-- At this point I'm ready to integrate it into the engine for CPU parallel read walkers
– Write general NanoScheduler framework in utils.threading. Test with reading via iterator from list of integers, map is int * 2, reduce is sum. Should be efficiency using resources to do sum of 2 * (sum(1 - X)).
Done!
CPU parallelism is nano threads. Pfor across read / map / reduce. Use work queue to implement.
Create general read map reduce framework in utils. Test parallelism independently before hooking up to Locus iterator
Represent explicitly the dependency graph. Scheduler should choose the work units that are ready for computation, that are marked as "completing a computation", and then finally that maximize the number of sequent available work units. May be worth measuring expected cost for read read / map / reduce unit and use it to balance the compute
As input is single threaded just need one thread to populate inputs, which runs as fast as possible on parallel pushing data to fixed size queue. Each push creates map job and links to upcoming reduce job.
Note that there's at most one thread for IO tasks, and all of the threads can contribute to CPU tasks
-- Invert logic in GATKArgumentCollection to disable monitoring, not enable. That means monitoring is on by default
-- Fix testing error in unit tests
-- Rename variables in ThreadAllocation to be clearer
-- Old version StateMonitoringThreadFactory refactored into base class ThreadEfficiencyMonitor and subclass EfficiencyMonitoringThreadFactory.
-- Base class is used by LinearMicroScheduler to monitor performance of GATK in single threaded mode
-- MicroScheduler now handles management of the efficiency monitor. Includes master thread in monitor, meaning that reduce is now included for both schedulers
-- Allows us to ID (by proxy) time spent doing IO
-- Refactor StateMonitoryingThreadFactory to use it's own enum, not Thread.State
-- Reliable unit tests across mac and unix
- Fix for M_Trieb's error report on the forum, and addition of integration tests to cover the walker.
- Addition of StructuralIndel as a class of variation within the VariantContext. These are for variants with a full alt allele that's >150bp in length.
- Adaptation of the MVLikelihoodRatio to work for a set of trios (takes the max over the trios of the MVLR)
- InsertSizeDistribution changed to use the new gatk report output (it was previously broken)
- RetrogeneDiscovery changed to be compatible with the new gatk report
- A maxIndelSize argument added to SelectVariants
- ByTranscriptEvaluator rewritten for cleanliness
- VariantRecalibrator modified to not exclude structural indels from recalibration if the mode is INDEL
- Documentation added to DepthOfCoverageIntegrationTest (no, don't yell at chartl ;_; )
Also sorry for the long commit history behind this that is the result of fixing merge conflicts. Because this *also* fixes a conflict (from git stash apply), for some reason I can't rebase all of them away. I'm pretty sure some of the commit notes say "this note isn't important because I'm going to rebase it anyway".
-- When merging multiple VCF records at a site, the combined VCF record has the QUAL of the first VCF record with a non-MISSING QUAL value. The previous behavior was to take the max QUAL, which resulted in sometime strange downstream confusion.
* No reads with Hard/Soft clips in the middle of the cigar
* No reads starting with deletions (with or without preceding clips)
* No reads ending in deletions (with or without follow-up clips)
* No reads that are fully hard or soft clipped
* No reads that have consecutive indels in the cigar (II, DD, ID or DI)
Also added systematic test for good cigars and iterative test for bad cigars.
-- Removed half-a*ssed attempt to automatically repair VCF files with bad headers, which allowed users to provide a replacement header overwriting the file's actually header on the fly. Not a good idea, really. Eric has promised to create a utility that walks through a VCF file and creates a meaningful header field based on the file's contents (if this ever becomes a priority)
-- All tests but one (using old bad VCF3 input) run unmodified with parallel code.
-- Disabled UNSAFE_VCF_PROCESSING for all but that test, which changes md5s because the output files have fixed headers
-- Minor optimizations to simpleMerge
-- BCF2 now determines whether it can safely write out raw genotype blocks, which is true in the case where the VCF header of the input is a complete, ordered subset of the output header. Added utilities to determine this and extensive unit tests (headerLinesAreOrderedConsistently)
-- Cleanup collapseStringList and exploreStringList for new unit tests of BCF2Utils. Fixed bug in edge case that never occurred in practice
-- VCFContigHeaderLine now provides its own key (VCFHeader.CONTIG_KEY) directly instead of requiring the user to provide it (and hoping its right)
-- More ways to access the data in VCFHeader
-- BCF2Writer uses a cache to avoid recomputing unnecessarily whether raw genotype blocks can be emitted directly into the output
-- Optimization of fullyDecodeAttributes -- attributes.size() is expensive and unnecessary. We just guess that on average we need ~10 elements for the attribute map
-- CombineVariants optimization -- filters are online HashSet but are sorted at the end by creating a TreeSet
-- makeCombinations is now makePermutations, and you can request to create the permutations with or without replacement
-- CombineVariants is now TreeReducible!
-- Integration tests running in parallel all pass except one (will fix) due to incorrect use of db=0 flag on input from old VCF format
-- Expanded unit tests
-- Support for clean logging of results to logger
-- Refactored MyTime into AutoFormattingTime in Utils, out of TraversalEngine, for cleanliness and reuse
-- Added docs and contracts to StateMonitoringThreadFactory
-- All low-level reads throw IOException instead of catching it directly. This allows us to not try/catch in readByte, improving performance by 5% or so
-- Optimize encodeTypeDescriptor with final variables. Avoid using Math.min instead do inline comparison
-- Inlined willOverflow directly in its single use
-- Old version converted doubles directly from strings. New version uses VariantContext getAttributeAsDouble() that looks at the values directly to determine how to convert from Object to Double (via Double.valueOf, (Double), or (Double)(Integer)).
-- getAttributeAsDouble() is now smart in converting integers to doubles as needed
-- Removed unnecessary logging info in BCF2Codec
-- Added integration tests to ensure that VQSR works end-to-end with BCF2 using sites version of the file khalid sent to me
-- Added vqsr.bcf_test.snps.unfiltered.bcf file for this integration test
-- VisualizeContextTree now can write out an equivalent BQSR table determined after adaptive context merging of all RG x QUAL x CONTEXT trees
-- Docs, algorithm descriptions, etc so that it makes sense what's going on
-- VisualizeContextTree should really be simplified when into a single tool that just visualize the trees when / if we decide to make adaptive contexts standard part of BQSR
-- Misc. cleaning, organization of the code (recalibation tests were in private but corresponding actual files were public)
-- We are no likely to fail with an error when reading old BCF files, rather than just giving bad results
-- Added new class BCFVersion that consolidates all of the version management of BCF
-- Previous version would count all alt alleles as present in a sample, even if only 1 were present, because of the way VariantEval subsetted VCs
-- Updated code for subsetting VCs by sample to be clearer about how it handles rederiving alleles
-- Update a few pieces of code to get previous correct behavior
-- Updated a few MD5s as now ref calls at sites in dbSNP are counted as having a comp sites, and therefore show up in known sites when Novelty strat is on (which I think is correct)
-- Walkers that used old subsetting function with true are now using clearer version that does rederive alleles by default
-- Moved most of BQSR classes (which are used throughout the codebase) to utils.recalibration. It's better in my opinion to keep commonly used code in utils, and only specialized code in walkers. As code becomes embedded throughout GATK its should be refactored to live in utils
-- Removed unncessary imports of BQSR in VQSR v3
-- Now ready to refactor QualQuantizer and unit test into a subclass of RecalDatum, refactor unit tests into RecalDatum unit tests, and generalize into hierarchical recal datum that can be used in QualQuantizer and the analysis of adaptive context covariate
-- Update PluginManager to sort the plugins and interfaces. This allows us to have a deterministic order in which the plugin classes come back, which caused BQSR integration tests to temporarily change because I moved my classes around a bit.
-- Check if a traversal error occurred in the last shard
-- Catch ExecutionException from the TreeReducer and throw as our HMS execption
-- ShardTraverser just throws the exception as formatted by the HMS, rather than wrapping it as a RuntimeException itself
-- EngineFeaturesIntegrationTests now uses public exampleFASTA (faster), and does 1000x iterations (slower)
-- Better error message when a traveral error occurs (a real bug)
-- EngineFeaturesIntegrationTest runs the multi-threaded error testing routines 50x times
-- A bit of cleanup in WalkerTest
-- VariantFiltration now properly sets passFilters in VC
-- BCF2 writer now properly decodes lazy BCF genotype data that it uses. Improper use generated a horrible subtle bug but the good news is that the extra checks I put in (unnecessarily a few days ago) caught the bug!
Signed-off-by: Mark DePristo <depristo@broadinstitute.org>
* Did not touch archived walkers... those can be named whatever.
* Kept abstract classes that end in Walker untouched (e.g. LocusWalker, ReadWalker, ...)
* Renamed a few inner classes due to conflict when stripping off Walker from their outer classes: ContigStats, FlagStats and FastaStats.
-- Heng wants to use 0x0? to represent any missing type value, which in our implementation was invalid. Updated our codebase to support this construct. Heng said he'll update the BCF2 quick reference.
-- Enabled integration test reading Heng's ex2.bcf file
-- GATK now only warns in the case where the END info field isn't the same (or +1 due to padding) as the getEnd() function as determined by the GATK. Turns out there's a single record in the 1000G SV call set that doesn't have the right length
-- VariantContextTestProvider now tests that X = Y where X -> writing -> reading -> writing -> reading = Y for a variety of variant context inputs X
-- Added integration test reading 1000G SV chr1 calls (from Chris)
-- If eval has genotypes and comp has genotypes, then subset the genotypes of comp down to the samples being evaluated when considering TP, FP, FN, TN status. This is important in the case where you want to use this to assess, for example, the quality of calls on NA12878 but you have a CEU trio comp VCF. The previous version was counting sites polymorphic in mom against the calls in NA12878.
-- Added testdata VCF and integrationtests to ensure this behavior continues in the future
-- TODO: actually run integration tests when I have an internet connection
-- If eval has genotypes and comp has genotypes, then subset the genotypes of comp down to the samples being evaluated when considering TP, FP, FN, TN status. This is important in the case where you want to use this to assess, for example, the quality of calls on NA12878 but you have a CEU trio comp VCF. The previous version was counting sites polymorphic in mom against the calls in NA12878.
-- Added testdata VCF and integrationtests to ensure this behavior continues in the future
-- getMetaData now split into getMetaDataInSortedOrder() [old functionality] and getMetaDataInOriginalOrder() [according to the header order]. Important as BCF uses the order of elements in the header in the offsets to keys, and we were automatically sorting the BCF2 header which is out of order in samtools and the whole system was going crazy
-- Updating GATK code to use the appropriate header function (this is why so many files have changed)
-- BCF2 code was busted in not differentiating PASS from . from FILTER in VC (tests coming that will actually stress this)
-- Bugfix for adding contig lines to BCF2 header dictionary
-- VCFHeader metaData no longer sorted internally. The system now maintains the data in header order, and only sorts output as requested in API
-- VCFWriter and BCF2Writer now explictly sort their header lines
-- Don't allow filters to be added that are PASS in the contract
with the new clipping behavior for weird cigars, we no longer can assert the final number of bases in the unit test, so I'm taking this bit off the unit test.
Eric Banks
Christopher Hartl
Mark DePristo
David Roazen
Menachem Fromer
Ryan Poplin
Guillermo del Angel
Mauricio Carneiro
Ami Levy Moonshine
Joel Thibault
Laurent Francioli
Khalid Shakir