Hello GATK team,
I'm trying to do some benchmarking with Mutect2 and a synthetic dataset. Even if the results are very good in terms of precision (~99%), recall or sensitivity can be better (~80%). In order to enhance sensitivity, i try to investigate why some variants are not called and in particular this variant :
chr1:201030579:A>T
I guess, it's because this variant is too near from the end of the reads.
So i tried to add "--pir_median_threshold 0 --pir_mad_threshold 0" but the result is the same.
Do you have any idea about how to call this kind of variant ?
this synthetic dataset is used to simulate amplicon reads.
java/jre1.8.0_76/bin/java -XX:GCTimeLimit=50 -XX:GCHeapFreeLimit=10 -Xmx16G -jar GenomeAnalysisTK-3.7/GenomeAnalysisTK.jar -T MuTect2 -R hg19.fa -I:synthetic_tumor.sort.region1.bam -I:synthetic_normal.sort.region1.bam --cosmic CosmicCodingMutsV75.hg19.sort.vcf -L test.bed -o mutect2.test.vcf -bamout bamout.bam -log mutect2.log
you will find enclosed :
the IGV snapshot
the tumor bam file in a zip file
the normal bam file in a zip file
the bed file in a zip file
Thank you for your help
