[ensembl-dev] Question Regarding VEP script warnings and --html option

[Will McLaren]

Hi Mathias,

The error is occurring because there's something odd about your VCF file, I
think.

The first warning, ($data[4] is uninitialized), happens because the 5th
column (the alternate allele, ALT) of your VCF line is not present - this
usually means that there are fewer than 5 columns of data on that
particular line, or the delimiter between the columns (space or tab) is not
present or corrupted somehow.

VEP expects at least the first 5 columns of the VCF (CHR, POS, ID, REF,
ALT) to be populated or to have a null "." value - have a check of your VCF
file to make sure that it looks OK. The vcftools package has a tool called
vcf-validator which you can use to validate the format of VCF files, see
http://vcftools.sourceforge.net/docs.html

Currently it's not possible to produce HTML (--html) and VCF (--vcf) output
- to see the HTML output, remove --vcf from your command line. If you want
the VCF output as well, you'd have to do that in a separate VEP run. If you
intended to indicate that the input format was VCF, you should use
"--format vcf".

Regards

Will McLaren
Ensembl Variation




On 11 December 2013 11:14, Mathias Lesche <
mathias.lesche at biotec.tu-dresden.de> wrote:

> Dear VEP developers,
>
> I have two questions regarding the current VEP script (Version 74).
>
> I call the script in the following way:
> /opt/rocks/bin/perl /projects/seq-work/user/all/vep/vep-74/variant_effect_
> predictor.pl -i input.vcf -o ouput.vcf --species homo_sapiens --fork 6
> --format vcf --cache --dir_cache /projects/seq-work/user/all/vep/vep-cache/
> --sift p --polyphen p --humdiv --total_length --numbers --protein
> --canonical --pubmed --gmaf --maf_1kg --maf_esp --vcf --html --offline
>
> When running the script I get the following warning:
> Use of uninitialized value $data[4] in string eq at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 415, <GEN0> line 837.
> Use of uninitialized value $ref in length at /projects/seq-work/user/all/
> vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm line 448, <GEN0> line 837.
> Use of uninitialized value $alt in pattern match (m//) at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 451, <GEN0> line 837.
> Use of uninitialized value $alt in split at /projects/seq-work/user/all/
> vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm line 540, <GEN0> line 837.
> Use of uninitialized value $alt in pattern match (m//) at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 549, <GEN0> line 837.
> Use of uninitialized value $ref in concatenation (.) or string at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 621, <GEN0> line 837.
> Use of uninitialized value $alt in concatenation (.) or string at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 621, <GEN0> line 837.
> Use of uninitialized value $data[2] in string eq at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 621, <GEN0> line 837.
> Use of uninitialized value in pattern match (m//) at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 2395, <GEN0> line 837.
> Use of uninitialized value in substitution (s///) at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 2395, <GEN0> line 837.
> Use of uninitialized value in string eq at /projects/seq-work/user/all/
> vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm line 2396, <GEN0> line 837.
> Use of uninitialized value in pattern match (m//) at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 2401, <GEN0> line 837.
> Use of uninitialized value in concatenation (.) or string at
> /projects/seq-work/user/all/vep/vep-74/Bio/EnsEMBL/Variation/Utils/VEP.pm
> line 2402, <GEN0> line 837.
>
> But the calculation of the variations finishes. I'm just wondering if it
> comes from the vcf file or is from the script itself.
>
> Secondly the output.vcf.html is empty. It doesn't show the variants or any
> links to Ensembl. Is there a way to fix it? But I use the --html option.
>
> I have to say I don't run vep on the whole vcf file. It is only a portion
> of it including the header.
>
> Regards,
> Mathias Lesche
> --
> Mathias Lesche, Dipl. Bioinformatician
> Deep Sequencing Group - SFB655
> c/o DFG-Center for Regenerative Therapies Dresden
> Cluster of Excellence / TU Dresden
> Fetscherstraße 105
> 01307 Dresden
>
> Phone: +49 (351) 458 82362
> Email: mathias.lesche(at)biotec.tu-dresden.de
>
> _______________________________________________
> 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/20131211/c4e695d0/attachment.html>