Bugs & Feature Requests
Known issues and requested enhancements

This page lists the issues tagged "bug" in the GATK4 development repository. If you have reported a bug and cannot find it in this list, let us know in the thread where you reported it. Note that we are no longer listing or processing issues in older versions of GATK (3.x and prior).

Showing open

GATK failing in FireCloud with google ExecutionException (caused by UnknownHostException)
User gets StackOverflowError when using Multi-interval in GenomicsDBImport with GATK
GATK4 VariantAnnotator does not handle PED file that GATK3 can
GenotypeGVCFs hangs in
Debug state is not properly passed to the EventMap in HaplotypeCaller
Mutect2 and haplotypecaller require UCSC chromosome names
Inconsistent unmapped but placed query results
Create a MAFCodec for eliminating possible conflicts
SV breakpoints location off when surrounding homology are of different length, and SVLEN off when duplicated copies are of different length
Malformed VCF headers in HaplotypeCallerSpark
The combination of ReadsDownsamplingIterator + ReservoirDownsampler does not work correctly
GenomicsDB errors discovered in the wild
Picard version doesn't print correctly when running a Picard tool from GATK
java.lang.IllegalArgumentException: Invalid interval.
barclay fails to set collection arguments initialized with an immutable collection
Please fix the broken 3.8-1 release with wrong pom.xml
BwaSpark and bam order
Users reporting unreasonable memory usage in GenotypeGVCFs
`__pthread_tpp_change_priority` error in GenotypeGVCFs (GATK
GATK4 GenomicsDB error with GenomicsDB versions 0.9.0 and 0.9.2
gatk sets TMP_DIR globally readable/writable for any user
GenotypeGVCFs takes a long time on small interval
funcotator and file locking
Error running HaplotypeCallerSpark in GVCF mode
Funcotator - Fix INDEL start/stop position and alleles for ALL output types
# in file names converted to %23 resulting in file not found
Spark tools have problems with umlauts
Funcotator - Problems with variant alleles that span exon start/end boundaries
MannWhitneyU shouldn't do a get(double)
Problem with spark.yarn.executor.memoryOverhead on DataProc