[ensembl-dev] upstream/downstream changes based on context in vcf?

[Andrew Parton]

Hi Matt,

Thanks for these, we’ve taken a quick look this morning and we can reproduce the issue. It looks like this is a problem specific to r89 of VEP, updating VEP to the latest version (95) should resolve this issue.

Kind Regards,
Andrew

> On 14 Jan 2019, at 22:16, Matt Wood <matt.wood at codifiedgenomics.com> wrote:
> 
> I've managed to put together a small example with two variants that show the behavior. I've attached two vcfs.
> 
> The variant we're looking at is 22:35998183TTGTGTC>T.
> 
> I've found I can reproduce the bug with just the following options:
> 
> perl /space1/software/vep/ensembl-vep/vep --dir /vep/cache/ --offline --vcf --regulatory -i has_downstream.vcf -o has_downstream.vcf.vep
> 
> If you run it on no_downstream.vcf we're getting the following output:
> 
> 22      35998183        .       TTGTGTC T       120     PASS    CSQ=-|TF_binding_site_variant|MODIFIER|||MotifFeature|MA0058.2|||||||||||||-1||||Max:MA0058.2|1|N|,-|intergenic_variant|MODIFIER|||||||||||||||||||||||| GT:VR:RR:DP:GQ  0/1:10:10:20:.
> 
> But, if we run it on has_downstream.vcf we get additional downstream effects for the variant we're looking at:
> 
> 22      35998183        .       TTGTGTC T       120     PASS    CSQ=-|downstream_gene_variant|MODIFIER|MB|ENSG00000198125|Transcript|ENST00000359787|protein_coding|||||||||||4622|-1||HGNC|6915||||,-|downstream_gene_variant|MODIFIER|MB|ENSG00000198125|Transcript|ENST00000397326|protein_coding|||||||||||4622|-1||HGNC|6915||||,-|downstream_gene_variant|MODIFIER|MB|ENSG00000198125|Transcript|ENST00000397328|protein_coding|||||||||||4622|-1||HGNC|6915||||,-|downstream_gene_variant|MODIFIER|MB|ENSG00000198125|Transcript|ENST00000401702|protein_coding|||||||||||4622|-1||HGNC|6915||||,-|TF_binding_site_variant|MODIFIER|||MotifFeature|MA0058.2|||||||||||||-1||||Max:MA0058.2|1|N|      GT:VR:RR:DP:GQ  0/1:10:10:20:.
> 
> The only difference being the addition of another variant in the vcf.
> 
> Our version info:
> 
> Versions:
>   ensembl              : 89.df47f96
>   ensembl-funcgen      : 89.678099a
>   ensembl-io           : 89.feefbc2
>   ensembl-variation    : 89.af9aae5
>   ensembl-vep          : 89.7
> 
> Thanks for taking a look,
> Matt
> 
> 
> On Mon, Jan 14, 2019 at 8:15 AM Andrew Parton <aparton at ebi.ac.uk <mailto:aparton at ebi.ac.uk>> wrote:
> Hi Matt, 
> 
> Upstream/Downstream consequences should be reported on the occasions where a variant is found to be 5’/3’ of a gene - it shouldn’t depend on surrounding variants within the VCF. Could you please send me an example so that I can reproduce it and take a closer look?
> 
> Thanks,
> Andrew
> 
> > On 11 Jan 2019, at 17:12, Matt Wood <matt.wood at codifiedgenomics.com <mailto:matt.wood at codifiedgenomics.com>> wrote:
> > 
> > I was hoping to get some clarification on how the upstream_gene_variant and downstream_gene_variant consequences are determined.
> > 
> > We've found several cases where a variant will have downstream or upstream consequences when that variant appears in one VCF, but will be annotated without those consequences when in another VCF. 
> > 
> > Our best guess is that the context of surrounding variants in the VCF is changing whether those consequences are applied.
> > 
> > Is that how it is supposed to work? We're running version 89.7.
> > 
> > Matt
> > _______________________________________________
> > Dev mailing list    Dev at ensembl.org <mailto:Dev at ensembl.org>
> > Posting guidelines and subscribe/unsubscribe info: http://lists.ensembl.org/mailman/listinfo/dev <http://lists.ensembl.org/mailman/listinfo/dev>
> > Ensembl Blog: http://www.ensembl.info/ <http://www.ensembl.info/>
> 
> _______________________________________________
> Dev mailing list    Dev at ensembl.org <mailto:Dev at ensembl.org>
> Posting guidelines and subscribe/unsubscribe info: http://lists.ensembl.org/mailman/listinfo/dev <http://lists.ensembl.org/mailman/listinfo/dev>
> Ensembl Blog: http://www.ensembl.info/ <http://www.ensembl.info/>
> <has_downstream.vcf><no_downstream.vcf>_______________________________________________
> 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/20190115/a76ea452/attachment.html>