Thresholds to filter out variants from vcf file

Question

[abedkurdi10]

Hi all. I would like to ask what are thresholds used to filter out variants from VCF file ? if provided, are they used as standards or it depends ?

I appreciate any help. Thank you.

Answer 1

[WouterDeCoster]

For that see the following pages from GATK:

• http://gatkforums.broadinstitute.org/gatk/discussion/2806/howto-apply-hard-filters-to-a-call-set
• https://software.broadinstitute.org/gatk/guide/article?id=3225
• https://software.broadinstitute.org/gatk/guide/article?id=6925

Obviously there is no solution to all problems but these guidelines will hopefully put you on the right track.

[abedkurdi10]

what if I need to annotate them using vcftools, is there any thresholds to follow like MinDP, QUAL, GQ ?

[WouterDeCoster]

Have you read the pages I linked?

[abedkurdi10]

yes, but GATK has different attributes for filtration, they are based on calculations...

[WouterDeCoster]

I don't understand why you would need to annotate them using vcftools and can't use GATK filter guidelines, but sure, go ahead.

Answer 2

[Ron]

Some posts you may want to refer:

What are good settings for filtering VCF files?

How Do You Usually Filter Variant Calling Results?

Practice Of Filtering Vcf Files (From Gatk)

http://ecole-bioinfo-aviesan.sb-roscoff.fr/sites/ecole-bioinfo-aviesan.sb-roscoff.fr/files/files/FiltrageVariantNLapaluRoscoff2015.pdf