[ensembl-dev] VEP v82: Different handling when SVTYPE is defined in VCF INFO field

[Cyriac Kandoth]

Very helpful info. Thanks Will.

~C

On Tue, Nov 24, 2015 at 6:04 AM, Will McLaren <wm2 at ebi.ac.uk> wrote:

> Hi Cyriac,
>
> The SVTYPE info field flags to the VEP that it should treat the input as a
> structural variant.
>
> These are treated as different object types by the VEP. The main effect is
> that sequence is not considered as it not normally provided explicitly for
> SVs.
>
> If you can, you should only use SVTYPE to annotate "true" structural
> variants, i.e. those described as e.g. <DEL> or <TDUP> in the ALT field.
> For any that have explicit sequence in the REF and ALT fields, you should
> remove the SVTYPE field as this will lead to more precise annotation in the
> VEP for these positions.
>
> Regards
>
> Will McLaren
> Ensembl Variation
>
> On 23 November 2015 at 22:55, Cyriac Kandoth <kandoth at cbio.mskcc.org>
> wrote:
>
>> Hi,
>>
>> For the minimalist VCF and VEP command below, using mus_musculus and
>> GRCm38, the output VCF appears to skip reporting fields like HGVSc,
>> ALLELE_NUM, and EXON/INTRON positions. This does not happen if I remove
>> "SVTYPE=DEL" from the input VCF. Am I breaking VCF conventions by
>> specifying SVTYPE=DEL? Should it only be used for larger structural
>> variants?
>>
>> Thanks,
>> Cyriac
>>
>> --
>> *test.vcf contains:*
>> ##fileformat=VCFv4.2
>> #CHROM POS ID REF ALT QUAL FILTER INFO
>> 11 96283450 . TG T . PASS SVTYPE=DEL
>>
>> *VEP v82 command with local GRCm38 cache:*
>> perl variant_effect_predictor.pl --offline --species mus_musculus
>> --assembly GRCm38 --vcf --input_file test.vcf --output_file test.vep.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/
>>
>>
>
> _______________________________________________
> 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/20151124/3e5794d8/attachment.html>