Why is it possible to visualize a mutation at a specific locus in the BAM file in IGV, but cannot be called using mutect2?

[Joyjoyjoyc]

Hi, everyone! I'm trying to call mutation with Mutect2 with RNA-seq, and my scripts are given below. I simply use one sample as a test, with a prior knowledge that the mutation in ASXL1(c.1934dupG) can be detected with a pretty high VAF, and I can also see this mutation by using bam file in IGV, but I really wonder why my scripts can not call this mutation before filtering? Thank you so much!!!

gatk Mutect2
-R ${ref}.fa
-I ${sam}/${sam}.BQSR.bam
-O ${sam}/gatk/${sam}_withpon.vcf
--create-output-bam-index FALSE
--af-of-alleles-not-in-resource 0.0000025
--create-output-variant-index false
--germline-resource /home/cuiyiran/data/mtDNA_mutation/reference/somatic-hg38_af-only-gnomad.hg38.vcf
--panel-of-normals /home/cuiyiran/data/mtDNA_mutation/reference/somatic-hg38_1000g_pon.hg38.vcf

gatk FilterMutectCalls
-R ${ref}.fa
-V ${sam}/gatk/${sam}_withpon.vcf
--create-output-variant-index false
-O ${sam}/gatk/${sam}_withpon_fv.vcf

bcftools norm -m -both ${sam}/gatk/${sam}_withpon_fv.vcf | bcftools norm -m +both -f ${ref}.fa ${sam}/gatk/${sam}_withpon_fv.vcf -Ov -o ${sam}/gatk/${sam}_withpon_norm.vcf
####annotation
perl ~/miniconda3/envs/vep/bin/vcf2maf.pl
--input-vcf ${sam}/gatk/${sam}_withpon_norm.vcf
--output-maf ${sam}/gatk/${sam}_withpon_vep.maf
--vep-path ~/miniconda3/envs/vep/bin/
--vep-data $vepcache
--ncbi-build GRCh38
--cache-version=112
--ref-fasta ${ref}.fa
--tumor-id ${sam}