Filtering somatic variants (annotated VCF files)

Question

[g.zachwyc]

Hello! I am currently in the process of filtering gene variants in an annotated VCF file. The variants I have in the file are present in tumours samples, but not peripheral blood, so the assumption is that they are somatic mutations. I am not sure how to go on about filtering them. It is my first project using VCF files and I have already filtered the variants based on consequence and impact, but I am not sure what cut-off values I should use for QUAL, #ALT, COV, %Freq. Or if there are any other columns that I should look at in the first place. I am finding it hard to balance the filtering criteria so that I wouldn't filter out those variants that might actually be of interest to me. I would be grateful for any help. Many, thanks

Answer 1

[dthorbur]

There is no one rule for filtering variants. It really depends on the quality of your data, the species you are sampling, and the questions you are asking. So without more information, it's difficult to help you.

There are many ways to score variants, and here are a couple useful resources I used when I was first starting out:

https://gatk.broadinstitute.org/hc/en-us/articles/360035890471-Hard-filtering-germline-short-variants

https://gatk.broadinstitute.org/hc/en-us/articles/360035531112--How-to-Filter-variants-either-with-VQSR-or-by-hard-filtering