[ensembl-dev] VEP VCF custom annotation for gnomad coverage

[William McLaren]

Hi Nikolas,

Yes, VEP (as of release/88 and the switch to the ensembl-vep repo) now matches VCF entries based on the alleles present, see [1] and [2].

You may disable allele matching by using the “overlap” type instead of “exact”, but this will report entries where any part of the coordinates match between the input and the entry in the VCF (i.e. a SNP will give a match to a long deletion that overlaps it). Perhaps this may not be a concern if excluding data based on coverage is your aim.

Hope that helps

Will McLaren
Ensembl Variation

[1] : http://www.ensembl.org/info/docs/tools/vep/script/vep_custom.html#custom_options
[2] : https://github.com/Ensembl/ensembl-vep#differences-to-ensembl-tools-vep


On 27 June 2017 at 12:24:41, Nikolas Pontikos (n.pontikos at ucl.ac.uk) wrote:
Dear Developers,

I want to add custom annotation to VEP.
Specifically I want to add the gnomad coverage downloaded from here:
http://gnomad.broadinstitute.org/downloads

This is important because a colleague pointed out that, at the moment, it not possible to distinguish between not covered and covered but zero allele freq in gnomad.

My first attempt was to convert to VCF with REF and ALT set to 'X' as they should not be needed by tabix.  However this doesn't work.

Does VEP use the REF and ALT in the matching then?  I thought it was simply using positions with tabix.

For now, I have converted the per position coverage files to bed instead and that seems to work ok, although the output is not as nice.

Many Thanks,

Nikolas.






_______________________________________________  
Dev mailing list Dev at ensembl.org  
Posting guidelines and subscribe/unsubscribe info: http://lists.ensembl.org/mailman/listinfo/dev  
Ensembl Blog: http://www.ensembl.info/  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20170627/ae61a0bc/attachment.html>