Parameter to filter a VCF of SNPs?

Question

[Picasa]

Hi,

I have done a SNPs analysis and now I have a VCF.

I would like to filter it but I don't know what is usually done. For instance I know that I can filter by:

• Remove samples that have > x% missing data
• Remove SNPs that have > x% missing data
• Remove SNPs with >x% MAF

Do you have have other recommendation?

And also what are the standard parameters (x %) ?

Thanks for your help.

Answer 1

[Pierre Lindenbaum]

Remove samples that have > x% missing data

search for the bad samples How to check which samples has more uncalled genotypes in multi-sample vcf and remove them with bcftools

Remove SNPs that have > x% missing data

gatk select variants https://software.broadinstitute.org/gatk/documentation/tooldocs/3.8-0/org_broadinstitute_gatk_tools_walkers_variantutils_SelectVariants.php with option --maxNOCALLfraction

Remove SNPs with >x% MAF

gatk select variants https://gatkforums.broadinstitute.org/gatk/discussion/5567/selectvariants-on-af