Mutect2
Can you please provide
a) GATK version used : 4.1.4.1
b) Exact GATK commands used: gatk Mutect2 -R $ref -I tumor.bam -O unfiltered.vcf
In this article (Mutect2)
gatk Mutect2 -R $ref -I tumor.bam -O unfiltered.vcf
this command does not include normal.bam. is it right command for somatic mutation?
i don't have panel of normal. for making panel of normal there are some recommendation like no tumor, blood, at least 40 samples number.
How can i do Mutect2?
-
9521ljh A matched normal should always be used if one is available, but Mutect2 can perform tumor-only calling if necessary. It is usually unnecessary to make your own panel of normals (the only exceptions would be non-human samples or samples with esoteric artifacts). Instead, use our public panels: gs://gatk-best-practices/somatic-hg38/1000g_pon.hg38.vcf.gz for hg38 and gs://gatk-best-practices/somatic-b37/Mutect2-exome-panel.vcf or gs://gatk-best-practices/somatic-b37/Mutect2-WGS-panel-b37.vcf for hg19/b37.
-
Hi Sir
I have currently 23 Tumor vs Normal samples. According to GATK Best Practices, it is desirable if not necessary to include PONS as described here and here. My question is would PONS using 23 Adjacent Tissue Normals suffice? Or should I combine these 23 with the 1000g_pon.hg38.vcf.gz to have a larger set?
Also, what's the source of the samples from which 1000g_pon.hg38.vcf.gz is created? Were these Normal Blood Samples or Adjacent Tissue Normals. Since I have Adjacent Tissue Normals, would it be principally correct to merge the blood normals and adjacent tissue normals?
Additionally, when I have more pairs available (say 20 more) maybe after few more months, do I need to provide each PON file separately that was used to generate the 23_combined_PONs.vcf.gz along with the 20 new PONS or can I simply use 23_combined_PONs.vcf.gz and other 20 PONS to get updated PONS?
-
Hi rohit satyam , here is the updated link for the Panel of Normals documentation.
You can use your own normals to make a PON, but it is usually only recommended if you have more than 40 samples. It is always better to have a PON from your own sequencing because the purpose is to find sequencing artifacts. However, you can use the 1000g_pon.hg38.vcf.gz file because sequencing machines generally have the same systematic errors. The 1000g_pon.hg38.vcf.gz is made from the 1000 genomes project, from blood samples of healthy people. I would not expect combining your 23 normals with the 1000 genomes project to make a large difference. What you decide depends on your specific case.
Please sign in to leave a comment.
3 comments