IllegalStateException in GenotypeGVCFs after GenomicsDBImport - GATK 4.2.6.1 #7938
Comments
|
Just reiterating here what @lbergelson noted in office hours: looks like the offending check was added in #7738, which ultimately affects both the ExcessHet and InbreedingCoeff annotations. @droazen reviewed that PR and might have more insight as to the desired behavior for these annotations when we are missing PLs due to GenomicsDB dropping them upstream---should we just not emit these annotations? |
|
@AJDCiarla The user should try re-running |
|
@AJDCiarla It would also be useful to know whether the error occurs when the user runs |
|
@droazen, like Karina posted in #7933, with our inputs this issue only occurs when using --force-output-intervals. I tried increasing --max-alternate-alleles to 2048 with no change. I just got back from a vacation, but this week I will try to debug this more closely to see what is causing the issue. Have you had any further discussions beyond what @samuelklee suggested above? |
* Support direct CRAM conversion in alignment pipelines * Remove jboss and standardize gradle files with develop * Switch error to warning * Update GenotypeGVCFHandler to include non-variant sites when making sites-only VCF * Update case in toLower * More informative error message * Fix argument in SamtoolsCramConverter * Update jbrowse dependencies (#165) * Add CRAM to allowable JBrowse track types * Initial support for pbmm2 and pbsv * Initial support for vulcan long read aligner * Support for quality metrics from nimble * Ensure output directory exists * Fix to JBrowse 2 CRAM tracks * Add additional vulcan alignment outputs * Fix filepath typo * Update nimble alignment defaults * Add debug message for nimble * Bugfix to nimble metrics import when running as alignment * Improve column width for nimble panels * Make nimble max_hits_to_report configurable * Skip merge unaligned for long-read aligners * Add UCell calculation step * Better handling for job resume after nimble failure * Add validation and bugfix for Nimble metrics import * Improve log messages for nimble metrics import * Support maxGenotypeCount for GenotypeGVCFs * Bump terser from 5.12.1 to 5.14.2 in /jbrowse (#166) Bumps [terser](https://github.com/terser/terser) from 5.12.1 to 5.14.2. - [Release notes](https://github.com/terser/terser/releases) - [Changelog](https://github.com/terser/terser/blob/master/CHANGELOG.md) - [Commits](https://github.com/terser/terser/commits) --- updated-dependencies: - dependency-name: terser dependency-type: indirect ... Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Bugfix to the order of nimble gz and error checks * Bugfix to the order of nimble gz and error checks * Add debugging * Add debugging to cDNA prep * Add view to assist in management of single-cell data * Add another view to assist in management of single-cell data * Remove no longer needed install of jbrowse 1 * Update report filter * Allow batch assignment of sequence libraries to runs by lane * Add new param to fail 10x processing if too few cells are found * Support nimble strandedness filter * Remove no-longer-needed nimble handler * Support CellMembrane and Seurat IntegrateData * Update SplitSeurat minCellsToKeep to allow fractions * Register new IntegrateData step * Support RIRA CalculateGeneComponentScores * Bugfix to RIRA CalculateGeneComponentScores * Skip GATK annotations to avoid broadinstitute/gatk#7938 * Return to using MS- prefix for cell hashing dual-index barcodes * Add additional single-cell filters and support ReblockGVCF * Allow celltypist model to run across genomes * Fix nextclade syntax * Fix nextclade syntax * Fix nextclade syntax * Fix nextclade syntax * Support GATK ReblockGVCF * Switch queries to use POST * Bugfix to ReblockGvcfHandler * Add admin action to manually update URI on ExpData objects * Allow GenotypeGVCFHandler to create genomicsdb workspaces on-the-fly * Ensure file is cached for GenotypeGVCFs exclude_intervals * More specific regex * Serialize SequenceAnalysisJobSupport outside of PipelineJob to reduce the size of the job's JSON * Allow deserialization of legacy JSON files containing support property * Update artifactory URLs * Improve unit test * Debug pipeline job serialization (#167) * Debug pipeline job serialization * Fix bug with gene scores not being saved * Increase RAM for remote FASTQC jobs * Test fixes * Only serialize SequenceJobSupport to disk when running on webserver * More fixes around serialization of SequenceJobSupport * Reduce HaplotypeCaller max-alternate-alleles * Support additional GenotypeGVCF params * Fix tests * Fix value for createsSeuratObjects on several steps * Improve warning messages * Improve warning messages * Improve warning messages * Improve GenotypeGVCFs logging * Improve GenotypeGVCFs logging and drop old params * Prior to GenotypeGVCFs, create workspaces with padding over the provided intervals * Refactor VcfComparisonStep to support VCF output(s), and add mGAP-release-specific version * Add validation for SamtoolsCramConverter * Reduce the amount of serialization to disk from SequenceJobSupport * Prepare sequence pipeline client code for non-savable params * Allow Seurat merge object name to be excluded from saved templates * Allow GenotypeGVCFs to locally cache support files * Update picard syntax to match upcoming argument changes * Further reduce sequence support serialization * Add creation of bgzipped genomes to standard genome import * Bugfix to genome gzipping * Update picard version for tests * Bugfix to alignment and skipping merge unaligned reads * Match picard version to sequence tests * Correct picard version * Improve logging Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Hello, did you deal with this probelm, I also encounter this. |
|
The error comes from two annotations: InbreedingCoeff and ExcessHet. One solution is to add "-AX ExcessHet -AX InbreedingCoeff". It doesnt exactly solve the problem, but it avoids hitting the problem code. |
Awesome! It is useful. Thank you very much! |
IllegalStateException in GenotypeGVCFs after GenomicsDBImport - GATK 4.2.6.1
Looks like there are similar issues occurring in #7639 and #7933. This is a follow up report from the GATK Forum.
GATK Forum Post: (https://gatk.broadinstitute.org/hc/en-us/community/posts/6972994559643-java-lang-IllegalStateException-in-GenotypeGVCFs-after-GenomicsDBImport-GATK-4-2-6-1)
Bug Report
Tools/Methods
GenotypeGVCFs --> GenomicsDBImport
Affected version(s)
-GenomicsDBImport: GATK 4.2.4.0
-GenotypeGVCFs: GATK 4.2.6.1
Description
IllegalStateException being thrown in GenotypeGVCFs after GenomicsDBImport. Exception denotes that "genome has no likelihoods". User is dividing into 50 intervals.
Stacktrace:
Exact Commands Used:
GenomicsDBImport:
java -Dsamjdk.use_async_io_read_samtools=false -Dsamjdk.use_async_io_write_samtools=true -Dsamjdk.use_async_io_write_tribble=false -Dsamjdk.compression_level=2 -Xms2G -Xmx20G -XX:+UseParallelGC -XX:ParallelGCThreads=2 -jar MySoftwares/gatk-4.2.6.1/gatk-package-4.2.6.1-local.jar GenomicsDBImport --genomicsdb-workspace-path 007_Database_DBImport_VCFref/database_interval_9 --sample-name-map sample_name_map --intervals 006_IntervalsSplit_DBImport_VCFref/interval_9.list --reader-threads 5 --batch-size 60 --tmp-dir TMPDIR --max-num-intervals-to-import-in-parallel 3 --merge-input-intervalsGenotypeGVCFs:
java -Dsamjdk.use_async_io_read_samtools=false -Dsamjdk.use_async_io_write_samtools=true -Dsamjdk.use_async_io_write_tribble=false -Dsamjdk.compression_level=2 -Xms4G -Xmx16G -XX:+UseParallelGC -XX:ParallelGCThreads=2 -jar MySoftwares/gatk-4.2.6.1/gatk-package-4.2.6.1-local.jar GenotypeGVCFs -R PigeonBatch5/000_DataLinks/000_RefSeq/Cliv2.1_genomic.fasta --intervals 006_IntervalsSplit_DBImport_VCFref/interval_9.list --force-output-intervals PigeonBatch4/008_RawVcfGz/MergeVcf/pigeonBatch1234_filtered.vcf.gz -V gendb://007_Database_DBImport_VCFref/database_interval_9 -O 008_RawVcfGz_DBImport_VCFref/001_DividedIntervals/interval_9.vcf.gz --tmp-dir TMPDIR --allow-old-rms-mapping-quality-annotation-data --only-output-calls-starting-in-intervals --verbosity ERRORUser Description of the Issue:
"I'm using the GenotypeGVCFs function based on GenomicsDBImport database. I've divided the reference into 50 intervals. Some intervals seems ok, but some reports error as following.
I used a VCF file in "--force-output-intervals" for down stream analysis. I've never seen this error without "--force-output-intervals". I've searched for the error message and changed my GATK version to 4.2.6.1 since similar error has been solved as a bug in recent update, but it still not works on my dataset..."
@droazen and @samuelklee , any insight on this?
Thank you,
Anthony
The text was updated successfully, but these errors were encountered: