Hi, What is the current option for "--cosmic" in Mutect2 tool name under GATK 4.0 beta version?
I am trying to call variants in a panel of unmatched normals.
There is no use of Cosmic in GATK4. Have a look at the latest presentations and tutorials on Mutect2. They are in the presentations section.
Also, have a look at this thread.
I was able to run Mutect2 using the below command.
./gatk-launch --javaOptions -Xmx12G Mutect2 -R $ref_fasta -I normal3.bam -D $DBSNP -tumor Normal3 -O output.normal3.vcf
I am trying to a create a PON; however, the output vcf file is completely empty. All reads seems to be filtered out; for instance:
"Mutect2 - 43410531 read(s) filtered by: (((((((((((MappingQualityReadFilter AND MappingQualityAvailableReadFilter) AND MappingQualityNotZeroReadFilter) AND MappedReadFilter) AND PrimaryAlignmentReadFilter) AND NotDuplicateReadFilter) AND PassesVendorQualityCheckReadFilter) AND NonZeroReferenceLengthAlignmentReadFilter) AND ) AND MateOnSameContigOrNoMappedMateReadFilter) AND GoodCigarReadFilter) AND WellformedReadFilter)"
Help please? Thank you
So, you are saying the output of Mutect2 on just normal samples is empty? If so, how did you pre-process your BAM files and what kind of data are you working with? Please also post the exact log output from the Mutect2 run.
Yes. I have generated the alignment files using STAR aligner, marked the duplicates using picard, addOrReplaceReadGroups, and then sorted the bam files. I am using transcriptome data.
Hmm. Looking at the filter status, it looks like your reads are badly mapped. Can you post some IGV screenshots of sites where you think variants should be called? Do the reads have good mapping qualities/base qualities?
I fixed the problem. It is the nature of the data type i.e, transcriptome data and STAR aligner --outSAMmapqUnique option. One of the previous post on GATK helped resolved the issue.
Wonderful news! If you can post the thread that ultimately helped you, it may help other users too