Mutation doesn't show up in VCF

[shmibble]

Hi,
I am doing an alignment of DNA from several target regions trying to find mutations. I know about one mutation that definitely exists in the sample I am running. It also shows up when looking at the sorted BAM file (12 times out of 450). However when I call the variants it does not show up in the resulting VCF file. I tried using BCFTools and Freebayes and none of them found it. With BCFTools I tried the -B -t DP and -t SP flags for the mpileup command. Does anyone have any suggestions on how to fix this?

Thanks!

[Bukowski]

Post your settings for FreeBayes/bcftools, otherwise we can't help you diagnose the problem. FreeBayes should be capable of detecting a mutation at 2% VAF with the right settings (dependent on the quality of the alternate alleles)

[shmibble]

Thanks for the reply.
I managed to find the mutation with FreeBayes using the -F flag (which filters out according to mutant allele frequency). In bcftools however I still can't seem to find it. I first run mpileup:

samtools mpileup -d8000 -B -t SP -t DP -m 1 -F 0.01 -Q 0 -go var.bcf -f ref.fa aligned_sorted.bam

Then:

bcftools call -vmO z -o var.vcf.gz var.bcf

Thanks again for the help.

[shmibble]

I should note that -B -t SP -t DP -m 1 -F 0.01 -Q 0 are all flags I tried recently hoping that they would help.

[Bukowski]

And what's the source of your data? Amplicon? in-solution capture? Have you deduplicated the data?

The Freebayes -F parameter defaults to 0.2 so did you reduce this to 0.01?

For low frequency alleles Freebayes will require also a minimum count of alternate alleles (2) - have you used the -m or -q parameters? Both of these will exclude low quality mapped reads, or bases from the genotyping.

Are your reads in the region mapped with high quality? What are the base qualities of your 12 alt allele observations?