how do you know an SV is located in exton, intron or Intergenic region

Question

[Maxine]

I have a VCF file including millions of structural variations. I want to filter those SVs by their locations (exons, introns, intergenic or mixed). Is any mature pipeline I can follow? Thanks in advance.

Maxine

[Pierre Lindenbaum]

get a BED of exons and the extract the SV using "bcftools view --regions-file exons.bed SV.indexed.vcf.gz"

[Maxine]

In my understanding, it would work to determine if the POS from VCF is located in an exon region. Am I right? But the thing is the POS only stores the start position of an SV, its end position is stored in INFO column.

[Pierre Lindenbaum]

the VCF should contain the INFO/END attrribute in the INFO column.

[Maxine]

Yes, the INFO/END exit in VCF file. But does "bcftools view --region-file ..." subset VCF based on only POS column? I wonder "bcftools view" take account of INFO/END.

[Pierre Lindenbaum]

I wonder "bcftools view" take account of INFO/END.

YES

and you can just try it.

[lieven.sterck]

you can try to run it through SnpEff ? (given that your genome is available for it)

[Maxine]

As I posted above, for SV, the end position should be considered. I'm not familiar with SnpEff, can it manipulate SV data?

[reza.jabal]

As lieven.sterck noted, your most straightforward option is SnpEff. If you have a standard VCF format, you don't need to manipulate anything. Do you have any idea what variant caller was used for calling the SVs?