GenotypeGVCFs runtime extremely long and not starting at 0%
a) GATK version used: 3.8-1
b) Exact GATK commands used
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=64
#SBATCH --ntasks-per-core=1
#SBATCH --time=240:00:00
#SBATCH --mem=300g
$JAVA $GATK \
-T GenotypeGVCFs \
-R $ref \
-V `find $outdir -name HaplotypeCaller.g.vcf | xargs | sed 's/ / -V /g'` \
-o $outdir/_calling/jointcalls.g.vcf \
-nt 64 \
($GATK refers to GenomeAnalysisTK.jar in my own directory)
c) The error log:
When I started running the job, it was kinda of bizarre that GenotypeGVCFs started on 2% instead of 0%. At the time the estimated running time is only 20h, which seemed normal to me as I was expecting finishing the job in less than a day using the multithreading option -nt 64:
INFO 17:24:48,596 ProgressMeter - chr1:63136701 0.0 24.9 m 2474.2 w 2.0% 20.4 h 20.0 h
However, after 20+h of running, the joint calling is still on chr1 and the estimated finish time kept increasing and increased to 5.8w:
INFO 13:27:35,331 ProgressMeter - chr1:63582801 0.0 20.5 h 15250.3 w 2.1% 5.9 w 5.8 w
Would appreciate if someone can let me know what went wrong. My colleague who used the same method didn't run into such a problem.
-
Hi,
We do not support GATK3 anymore. Please upgrade to the latest version of GATK4.
Please sign in to leave a comment.
1 comment