This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to use BCFtools to keep variants if they are `DP > 10` in all samples?

Dear Biostars,

To clarify this question is about best practices with trio-samples (proband, mother, father). I have been able to remove low confidence variants from .vcf.gz files via:

bcftools filter -Oz -i'FMT/DP>10' ${path}/x.vcf.gz > ${path}/x_filt.vcf.gz

This removes variants where the DP is less than 10 all three samples. However, this keeps variants where one sample has more than 10 reads, and the others do not. Below are two lines from x_filt.vcf.gz:

Format            Pband                  Mother                   Father
GT:AD:DP:GQ:PL    0/0:5,0:5:15:0,15,209  0/1:9,12:21:99:438,0,342 0/0:2,0:2:6:0,6,84
GT:AD:DP:GQ:PL    0/0:6,0:6:15:0,15,225  0/1:8,5:13:99:186,0,300  1/1:0,2:2:6:90,6,0

The DP is <10 in the Proband and Father for both variants, but passes the filtration due to the Mother having DP >10.

How can I remove cases such as these? Preferably using BCFtools, however other suggestions are welcome.

Many Thanks, Krutik

bcftools variant-calling genomics
• 1,963 views • link
updated by Ram Ram • written by K.patel5 • K.patel5

can you try filter by MIN(DP)>10

• 0 views • link
written by cpad0112 cpad0112

1 answer

(not tested)

 bcftools view -i 'MIN(FORMAT/DP)>10'
• 0 views • link
written by Pierre Lindenbaum Pierre Lindenbaum

Thanks, that did the trick.

• 0 views • link
written by K.patel5 • K.patel5

Log in to answer this question.

More posts like this