Variant Annotator not annotating MappingQualityRankSumTest and ReadPosRankSumTest for INDELs

TechnicalVault

Hi,

I'm attempting to use Variant Annotator to annotate some VCFs produced by samtools so I can run VQSR on them. Unfortunately I've gottent stuck and I'm trying to figure out why Variant Annotator wouldn't be annotating INDELs with MappingQualityRankSumTest and ReadPosRankSumTest, it seems to annotate SNPs fine. There are both Homs and het's called on the sample. Could it be I need to left align the indels to get enough coverage? What would you suggest is the best way to debug this? Is there a way to make GATK behave more verbosely about why it's refusing an annotation?

Thanks Martin

ebanks

You are using an older version of the GATK. This functionality was not implemented for indels until version 2.2.

ebanks

Ah, yes, I do see that now in the code. Many of these annotations can only be generated when going through the Unified Genotyper or Haplotype Caller. I'll make a note that we need to extend this functionality, but can't promise an ETA for that. In the meantime, perhaps you can run your calls through the UG in GENOTYPE_GIVEN_ALLELES mode? Sorry for the inconvenience.

Kurt

Not sure why they are not annotating, but in regards to making GATK more verbose, have you already added -l DEBUG to your command line?

TechnicalVault

Thanks for the suggestion, -l DEBUG does make matters more verbose, but it appears there's no debugging output from the annotator itself. I've now tried running left align then annotating, still flummoxed.

TechnicalVault

Hmm also they seem to all be coming out with FS=0.000;HaplotypeScore=0.0000.

If it helps this is the command line used:

java -Xmx2000m -Xms2000m -server -XX:+UseSerialGC -jar /nfs/users/nfs_m/mercury/src/GenomeAnalysisTK-1.6/GenomeAnalysisTK.jar -T VariantAnnotator -R /lustre/scratch111/resources/vrpipe/ref/Homo_sapiens/1000Genomes/human_g1k_v37.fasta -I /lustre/scratch111/projects/helic/vcf_intermediate-2/lists/chr1-pooled.list --variant /lustre/scratch111/projects/helic/vcf_intermediate-2/pooled/1/1:14000001-15000000.samtools.vcf.gz -L 1:14000001-15000000 -A QualByDepth -A HaplotypeScore -A MappingQualityRankSumTest -A ReadPosRankSumTest -A FisherStrand -A InbreedingCoeff -A DepthOfCoverage -o /lustre/scratch111/projects/helic/vcf_intermediate-2/pooled/1/1:14000001-15000000.vcf.gz.part.vcf.gz

TechnicalVault

My bad that was the command line from the older directory:

java -Xmx2500m -Xms2500m -server -XX:+UseSerialGC -jar /nfs/users/nfs_m/mercury/src/GenomeAnalysisTK-2.2/GenomeAnalysisTKLite.jar -T VariantAnnotator -R /lustre/scratch111/resources/vrpipe/ref/Homo_sapiens/1000Genomes/human_g1k_v37.fasta -I /lustre/scratch111/projects/helic/vcf_intermediate-3/lists/chr1-pooled.list --variant /lustre/scratch111/projects/helic/vcf_intermediate-3/pooled/1/1:14000001-15000000.samtools.vcf.gz -L 1:14000001-15000000 -U LENIENT_VCF_PROCESSING -A QualByDepth -A HaplotypeScore -A MappingQualityRankSumTest -A ReadPosRankSumTest -A FisherStrand -A InbreedingCoeff -A DepthOfCoverage -o /lustre/scratch111/projects/helic/vcf_intermediate-3/pooled/1/1:14000001-15000000.vcf.gz.part.vcf.gz

Rather than being 0 FS and HaplotypeScore are simply missing from the 2.2 annotated BAMs, similar to ReadPosRankSumTest and MappingQualityRankSumTest. Is the new functionality in the Lite GATK or just in Full?