How to slice a gvcf file with tabix?

Question

[Sd]

I am using the following tabix command to slice a GVCF file, but it is not creating sliced GVCF files. I have a bed file for the hg38 genome reference with 1MB regions and 1kb padding, and I want to slice using the bed file. I was wondering if you could help me with how to do it?

./tabix -R regions.bed -p vcf sample.g.vcf.gz > sample_sliced.gz

Answer 1

[Pierre Lindenbaum]

why would you need to slice a gvcf file ? all the tools in gatk can use a BED / interval / chromosome to narrow the analysis

[Sd]

I want to run gatk GenomicsDBImport tool to build a GenomicsDB. I want to slice it to reduce the computation time since I have 1300 gvcf files. What is your suggestion to run GenomicsDBImport efficiently?

[Pierre Lindenbaum]

What is your suggestion to run GenomicsDBImport efficiently?

call the GVCF by slices eg: chr1:1000000-2000000
run GenomicsDBImport for each sample for one interval
genotype gvcf for one interval

at the end concatenate all the VCFs

[Sd]

Then what is the best way to do slicing for 1Mb windows? Doing --intervals in the GenomicsDBImport?

[Pierre Lindenbaum]

No, in haplotypecaller

[Sd]

I already have gvcfs called by chromosome and there are 24 gvcfs per sample which I merged them into a single gvcf per sample. Now I want to slice them to smaller chunks.