[ensembl-dev] VEP (ver78) TranscriptVariationAllele.pm unitialized value error with --hgvs enabled

Rocky Bernstein rocky.bernstein at gmail.com
Thu Jan 29 22:00:38 GMT 2015


Since the only use of a string is $hgvs_notation->{ref}, a guess is to
change line 1220 to:

  unless( defined($hgvs_notation->{ref}) && $hgvs_notation->{ref} eq "-" )

This masks the problem of why this didn't get defined in the first place
which might be a more serious matter.

On Thu, Jan 29, 2015 at 4:47 PM, Homann, Oliver <ohomann at amgen.com> wrote:

>  Hello,
>
>
>
> I have been using the VEP to annotate a GATK-generated VCF containing
> indels from a WGS genotyping project, and have encountered a small number
> of entries that result in the following error (only when –hgvs is enabled):
>
>     “Use of uninitialized value in string eq at
> /<mypath>/ensembl-tools-release-78/scripts/variant_effect_predictor/Bio/EnsEMBL/Variation/TranscriptVariationAllele.pm
> line 1220”
>
>
>
> Here are the offending lines in the perl module:
>
>      1219   ### Convert peptide to 3 letter code as used in HGVS
>
>      1220   unless( $hgvs_notation->{ref} eq "-"){
>
>     1221       $hgvs_notation->{ref}  =
> Bio::SeqUtils->seq3(Bio::PrimarySeq->new(-seq => $hgvs_notation->{ref}, -id
> => 'ref',  -alphabet => 'protein')) || "";
>
>      1222   }
>
>
>
> I tracked down one of the offending entries, but it isn’t immediately
> obvious to me why it is causing an error:
>
>
>
> #CHROM             POS        ID            REF         ALT
> QUAL    FILTER   INFO      FORMAT              MySample
>
> 13           21729289             .               T
> TCAGTTTTCTTTGTTGCTGACATCTCGGATGTTCTGTCCATGTTTAAGGAACCTTTTA
> 118.19   .
> AC=1;AF=5.556e-03;AN=180;BaseQRankSum=0.744;DP=2652;FS=3.538;GQ_MEAN=66.68;GQ_STDDEV=20.39;InbreedingCoeff=-0.0085;MLEAC=1;MLEAF=5.556e-03;MQ=60.00;MQ0=0;MQRankSum=0.609;NCC=0;QD=0.13;ReadPosRankSum=-1.220e-01;SOR=0.252
> GT:AD:DP:GQ:PL              0/1:41,8:.:99:160,0,14558
>
>
>
> Any ideas what the problem might be?
>
>
>
> Thanks,
>
> Oliver Homann
>
>
>
> P.S.  Apart from this small glitch, VEP is teriffic!
>
> _______________________________________________
> 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/20150129/abd3d50d/attachment.html>


More information about the Dev mailing list