[ensembl-dev] Could not find variation cache for

[Will McLaren]

Hi Roland,

You can ignore that warning message; when you specify --everything, it
switches on a few options which tell the VEP to expect to find cache files
containing co-located variants. Since you generated your cache yourself,
these files don't exist, which is why the code is complaining. You can
either continue to ignore the warnings, or substitute --everything for the
list of flags specified here:

http://www.ensembl.org/info/docs/tools/vep/script/vep_options.html#opt_everything

In fact in your case only the following will work with a user-generated
cache anyway: --variant_class, --biotype, --numbers

Regarding the lack of protein-changing results, there is every chance that
the cache has not been generated correctly from the GTF. I notice you
converted a GFF; it's worth checking that the requirements on the input GTF
are quite strict, see
http://www.ensembl.org/info/docs/tools/vep/script/vep_options.html#opt_everything

It is on our to-do list to make this script compatible with a wider
spectrum of GFF/GTF formatting.

Regards

Will

On 5 June 2015 at 13:52, Schmucki, Roland <roland.schmucki at roche.com> wrote:

> Dear Will
>
> Thank you very much for the quick response.
> I would like to post this issue to the public Ensembl mailing list.
> Here is a brief description of the problem I encountered:
>
>
> When running VEP with ensembl annotation files I get errors of the form
> "Could not find variation cache for Chromosome..."
>
> I downloaded a  genome (i.e. pao1, $name.fa) and annotation ($name.gff3)
> from Ensembl ftp and then created the cache files according to the VEP
> tutorial:
>
>
> sort -k1,1 -k4,4n $name.gff | bgzip > $name.gff.gz
> tabix -p gff $name.gff.gz
> ./cufflinks/gffread $name.gff -T -o $name.gtf
> perl gtf2vep.pl -i $name.gtf -f $name.fa -d 79 -s $name --dir
> variant_effect_predictor_version79/cache_files_
> and move the cache files to the correct location manually.
>
> This all seem to have worked fine without any error or warning messages.
> Then I mapped the reads to the genome, ran Freebayes (variants.vcf with
> 2700 variants) and at the very end applied VEP with the following command:
>
>
> perl variant_effect_predictor.pl --everything --offline --custom
> $name.gff.gz,$name-genes,gff,overlap,0 --format vcf -i variants.vcf -o
> variants.txt --species $name --dir_cache $VEP_DATA
>
>
> The variable VEP_DATA points to the corresponding cache file:
> with the following files (creation date and file size) there in:
> $VEP_DATA/pao1/79/Chromosome/
> 292135 Jun  5 09:10 3000001-4000000.gz
> 294904 Jun  5 09:10 1000001-2000000.gz
> 290186 Jun  5 09:10 1-1000000.gz
> 290763 Jun  5 09:10 5000001-6000000.gz
> 284789 Jun  5 09:10 2000001-3000000.gz
> 292462 Jun  5 09:10 4000001-5000000.gz
> 78483 Jun  5 09:10 6000001-7000000.gz
>
>
> When I run VEP I get the following errors and warnings (See attached log
> file for all details):
> WARNING: Could not find variation cache for Chromosome:1-1000000
> WARNING: Could not find variation cache for Chromosome:5000001-6000000
> etc.
>
>
> I don't understand why I got this errors/warnings?
> Thanks a lot for any advice!
>
> Best,
>
> R.
>
>
> PS: there is an output file generated with variant annotations of the form:
>
> #Uploaded_variation     Location        Allele  Gene    Feature
> Feature_type    Consequence     cDNA_position   CDS_position    Pro
> tein_position        Amino_acids     Codons  Existing_variation      Extra
> Chromosome_2415_G/T     Chromosome:2415 T       gene:PA0005
> transcript:AAG03395     Transcript      downstream_gene_variant -
>        -       -       -       -       -
> IMPACT=MODIFIER;pao1-genes=gene:PA0002,exon_Chromosome:2056-3159,CDS:AAG03392,transc
>
> However, no amino acid changes are found which is unlikely.
>
>
> _______________________________________________
> 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/20150608/6d8dd8df/attachment.html>