Extract SNPs by position from gvcf

Question

[wangdavid758]

I have 10 gvcfs for 10 samples and a list of SNP positions (chr# and position) I want to extract from the gvcfs. I can't just use vcftools to do this because the position is given as a range (start position and end position) in the gvcf and if the position in my list is within the range but not the start of the range, it won't be extracted. How can I combine my 10 gvcfs into a single vcf with only the SNPs at the positions I specify in my list?

Answer 1

[Pierre Lindenbaum]

you don't do that. from a gvcf you call GenotypeGVCFs https://software.broadinstitute.org/gatk/documentation/tooldocs/3.8-0/org_broadinstitute_gatk_tools_walkers_variantutils_GenotypeGVCFs.php using the option -L for the desired position(s). ( https://software.broadinstitute.org/gatk/documentation/tooldocs/3.8-0/org_broadinstitute_gatk_engine_CommandLineGATK.php )

[rafal.gutaker]

What can one do to extract calls that are fixed for reference? They get lost after GenotypeGVCFs.

I need to add SNP calls from new dataset (outgroup species) to my old dataset (species of interest). They were mapped to the same genome and I called haplotypes using GATK. Extracting SNPs from gvcf is my only idea right now.

Many thanks!