Service Notice: Due to the blizzard currently hammering the US Northeast, the Broad is shut down and the GATK forum will be mostly unattended while we hunker down and sip hot cocoa with marshmallows. Assuming the power stays on and we're able to dig ourselves out of the snow when it's all over, normal service should resume Wednesday or Thursday.

HaplotypeCaller rs annotation

flescaiflescai Posts: 53Member ✭✭

Hi there, I'm running now the new GATK 2.2-2 version and I noticed an issue with HaplotypeCaller I had in the previous version I was using. Despite adding the dbSNP ROD to the walker, the emitted VCF doesn't contain rs names in the name field. On the contrary, UnifiedGenotyper annotates the variants with the appropriate names.

In my .scala code I wrote:

 class HaplotypeCallerArguments (t: Target) extends HaplotypeCaller with UNIVERSAL_GATK_ARGS {
   this.reference_sequence = qscript.referenceFile
   this.intervals = if (qscript.intervals == null) Nil else List(qscript.intervals)
   // Set the memory limit to 6 gigabytes on each command.
   this.memoryLimit = 6
   this.input_file :+= qscript.bamFile
   this.D = qscript.dbSNP_b37

and that is correctly reflected when queue launches the job as

 INFO  16:07:30,655 FunctionEdge - Starting:  'java'  '-Xmx6144m'  '-XX:+UseParallelOldGC'  '-XX:ParallelGCThreads=4'  '-XX:GCTimeLimit=50'  '-XX:GCHeapFreeLimit=10'  ''  '-cp' '/share/apps/genomics/Queue-2.2-2-gf44cc4e/Queue.jar'  
 'org.broadinstitute.sting.gatk.CommandLineGATK'  '-T' 'HaplotypeCaller'  '-I' '/SAN/biomed/analysis/recal.list'  '-L' '/SAN/biomed/analysis/.queue/scatterGather/HaplotypeCaller-sg/temp_016_of_300/scatter.intervals'  '-R' '/share/apps/genomics/reference/human_g1k_v37.fasta'  
 '-l' 'INFO'  '-o' '/SAN/biomed/analysis/.queue/scatterGather/HaplotypeCaller-sg/temp_016_of_300/comparisonHC.raw.vcf' '-D' '/share/apps/genomics/reference/gatkresources_hg19_1.5/'  

However, my VCF still looks like

grep -v \# HC.raw.vcf | cut -f 1,2,3,4,5 | more
1   762273  .   G   A
1   865738  .   A   G
1   866319  .   G   A
1   866511  .   C   CCCCT
1   871042  .   C   CA
1   874734  .   C   T

Am I doing something wrong? It would be quite time consuming to launch VariantAnnotation if not necessary, as I understand now the covariates used by VQSR are already emitted by the caller.

thanks, Francesco


Best Answer


Sign In or Register to comment.