Freebayes - SNP detection on 4 strains

[heathacoats]

Hi All,

I'm currently trying to detect variants across 4 different samples (each compared to my reference) using freebayes. I've attached read group tags to each of the samples so freebayes knows they are different, but my output is still only between the reference and 1 alternate (I assume freebayes is combining all 4 samples to represent the alternate case). Does anyone know how to run freebayes (or another variant calling program at that) so I have variant information for each of my 4 samples individually against my reference? I was hoping to do this in one run rather than 4 (one for each sample) - just to save myself some downstream analysis time.

Thanks!

Heather

[GARMALL]

Hi Heather,

I found myself in this problem. I solved it by using the bamaddgr utility (https://github.com/ekg/bamaddrg). The main issue is that the RG attached to the bam files must have a unique identifier. With this utility you can change the sample and the read group names to each bam file.
This would be the trick (I usually call variants per Mb):

bamaddrg --clear -R 1:0..1000300  -b BAM1.bam -s Name1 -r RG1 -b BAM2.bam -s Name2 -r RG2  | freebayes -r 1:0..1000000 --stdin -f FASTA.fa |bgzip > VCF.vcf.gz

Hope this helps.