Mutect2 Stall on Chromosome 3
REQUIRED for all errors and issues:
a) GATK version used:
GATK 4.5.0.0
b) Exact command used:
gatk --java-options "-Xmx36g" \
Mutect2 \
-R $my_reference \
-I Tumour.markdup.bam -I Normal.markdup.bam \
--tumor-sample Tumour \
-L chr3 -normal Normal \
-O Tumour.chr3.unfiltered.vcf.gz \
--f1r2-max-depth 500 \
--max-reads-per-alignment-start 0 \
--callable-depth 5 \
--f1r2-tar-gz Tumour.chr3.f1r2.tar.gz
Using GRCh38 reference genome.
c) Entire program log:
Just the bottom of the log file as it's been like this for 8 hours:
20:00:11.905 INFO ProgressMeter - chr3:87876867 93.8 381740 4070.2
20:00:22.031 INFO ProgressMeter - chr3:88026285 94.0 382380 4069.7
20:00:32.253 INFO ProgressMeter - chr3:88163106 94.1 382990 4068.8
20:00:42.387 INFO ProgressMeter - chr3:88306366 94.3 383650 4068.5
20:00:52.473 INFO ProgressMeter - chr3:88470270 94.5 384350 4068.7
20:01:02.533 INFO ProgressMeter - chr3:88609506 94.6 384970 4068.0
20:01:12.620 INFO ProgressMeter - chr3:88768639 94.8 385650 4068.0
20:01:22.752 INFO ProgressMeter - chr3:88951332 95.0 386430 4068.9
20:01:32.799 INFO ProgressMeter - chr3:89085527 95.1 387030 4068.1
20:01:42.878 INFO ProgressMeter - chr3:89238390 95.3 387700 4067.9
20:01:52.971 INFO ProgressMeter - chr3:89426714 95.5 388490 4069.0
20:02:03.089 INFO ProgressMeter - chr3:89592319 95.6 389200 4069.3
20:02:13.137 INFO ProgressMeter - chr3:89767884 95.8 389940 4069.9
20:02:23.264 INFO ProgressMeter - chr3:89938242 96.0 390670 4070.4
20:02:33.360 INFO ProgressMeter - chr3:90112413 96.1 391420 4071.0
20:02:43.627 INFO ProgressMeter - chr3:90263852 96.3 392080 4070.6
20:02:54.211 INFO ProgressMeter - chr3:90352492 96.5 392630 4068.9
20:03:04.701 INFO ProgressMeter - chr3:90418930 96.7 393070 4066.1
20:03:14.841 INFO ProgressMeter - chr3:90478156 96.8 393490 4063.3
20:03:24.872 INFO ProgressMeter - chr3:90518379 97.0 393750 4059.0
20:03:36.303 INFO ProgressMeter - chr3:90543437 97.2 393900 4052.6
20:03:46.396 INFO ProgressMeter - chr3:90591366 97.4 394190 4048.6
20:03:56.569 INFO ProgressMeter - chr3:90635005 97.5 394480 4044.5
20:04:06.590 INFO ProgressMeter - chr3:90672658 97.7 394730 4040.2
20:04:16.937 INFO ProgressMeter - chr3:90703397 97.9 394940 4035.2
20:04:27.004 INFO ProgressMeter - chr3:90832003 98.0 395610 4035.1
20:04:37.172 INFO ProgressMeter - chr3:90917939 98.2 396210 4034.3
20:04:47.180 INFO ProgressMeter - chr3:91085353 98.4 397310 4038.6
20:04:57.259 INFO ProgressMeter - chr3:91317440 98.5 398660 4045.4
20:05:07.332 INFO ProgressMeter - chr3:91486348 98.7 399370 4045.7
20:05:17.342 INFO ProgressMeter - chr3:91921344 98.9 401350 4058.9
20:05:27.350 INFO ProgressMeter - chr3:93059672 99.0 406300 4102.1
Seem's to progress fine, and then doesn't move once it hits chromosome 3 centromere region.
All other chromosomes finished Mutect2 within 270 minutes, but chromosome 3 hasn't finished after 16 hours.
-
Can you provide details of the system you are running this on? Did you check that region visually to see if there is anything unusual such as too high depth or heavy masking?
-
Hi,
I'm running on a Slurm scheduled high performance compute cluster.
My runtime parameters for this job were: 8 cores on a single node, 48 GB RAM, and 1TB of scratch space.I checked the bam file and there was nothing out of the ordinary in the region where the stall was or around it, other than it being the centromere of chromosome 3.
As mentioned all of the other chromosomes finished fine in a good time with the same runtime parameters.
-
Is it possible for you to split chr3 into different intervals split by hard masked (N) regions to try if you see any improvement? That way we may be able to pinpoint which region is causing this issue much better.
Also the region that GATK seems to stall is within a long Alpha Satellite Repeat region of chr3 which really is a long repeat region. You may be able to skip these regions in all chromosomes unless you have special interest in them.
-
Hi,
I made a change and used the --exclude-intervals option.
I gave it a file of centromere positions of which the chromosome 3 intervals are these:
chr3:91553419-93655574
chr3:90772458-91233586
chr3:91233686-91247622
Rerunning the command with this added in has allowed it to get past the sticky position already.
18:12:42.471 INFO ProgressMeter - chr3:132808803 154.0 566210 3677.4
18:12:52.548 INFO ProgressMeter - chr3:132922881 154.1 566700 3676.6
18:13:02.589 INFO ProgressMeter - chr3:133073861 154.3 567340 3676.7
18:13:12.609 INFO ProgressMeter - chr3:133215288 154.5 567960 3676.8
So it's something within one of those three regions. -
Great to hear that. It is possible that the repeat region could have accumulated too many reads therefore Mutect2 reassembly engine got stuck.
I hope it works this time.
Regards.
Please sign in to leave a comment.
5 comments