[ensembl-dev] Duplicates in VEP for Plugins and Custom flags

[Will McLaren]

Hi Kirsley,

You do not need to specify the config file manually with "--config
cache/vep.ini" if your --dir parameter is set to "cache".

The VEP looks for a config file called vep.ini in the cache directory
(default is $HOME/.vep but it looks like you've changed it to ./cache/) and
loads this automatically; you are then manually telling it to load the same
file again, so the configuration gets duplicated for things configured as
arrays i.e. plugins and custom files.

Either rename your ini file and specify it manually, or leave it as it is
and remove "--config cache/vep.ini" from your command.

Regards

Will

On 22 July 2015 at 22:01, Kirsley CHENNEN <kchennen at unistra.fr> wrote:

>  Hi Will,
>             There seems to be a bug in VEP (v81) for the use of Plugins
> and custom annotations. When theses parameters are specified in a config
> file, it seems that VEP imports them twice! This is even worse for custom
> annotations, as they are ouput twice and in a concatenated form! However,
> they behave normally, if the parameters are given in command line. Is it
> the only solution?
>
>
>  Plugin example:
> -------------------------
> * In config file:
> plugin               Blosum62
>
>
> * Shell log:
> [15-07-22 22:43:57] > ./variant_effect_predictor.pl --config
> cache/vep.ini -i example_GRCh37.vcf
> ...
> version 81
>
> By Will McLaren (wm2 at ebi.ac.uk)
>
> Configuration options:
>
>                    ##
> allele_number      1
> assembly           GRCh37
> config             cache/vep.ini
> core_type          core
> dir
> xxxxx/ensembl-tools/scripts/variant_effect_predictor/cache
> dir_plugins
> xxxxx/ensembl-tools/scripts/variant_effect_predictor/cache/Plugins
> force_overwrite    1
> fork               10
> host               ensembldb.ensembl.org
> input_file         example_GRCh37.vcf
> minimal            1
> offline            1
> phased             1
> plugin             Blosum62    Blosum62         <====
> port               3306
> species            homo_sapiens
> stats              HASH(0x42c7728)
> stats_text         1
> terms              SO
> variant_class      1
> verbose            1
>
> --------------------
>
> 2015-07-22 22:44:10 - Read existing cache info
> 2015-07-22 22:44:10 - Auto-detected FASTA file in cache directory
> 2015-07-22 22:44:10 - Checking/creating FASTA index
> 2015-07-22 22:44:10 - Loaded plugin: Blosum62   <====
> 2015-07-22 22:44:10 - Loaded plugin: Blosum62   <====
> 2015-07-22 22:44:10 - Starting...
> 2015-07-22 22:44:10 - Detected format of input file as vcf
> 2015-07-22 22:44:10 - Read 173 variants into buffer
> 2015-07-22 22:44:10 - Calculating consequences
> [===============================================]  [ 100% ]
> 2015-07-22 22:44:11 - Writing output
> 2015-07-22 22:44:11 - Processed 173 total variants (173 vars/sec, 173
> vars/sec total)
> 2015-07-22 22:44:11 - Wrote stats summary to
> variant_effect_output.txt_summary.txt
> 2015-07-22 22:44:11 - Finished!
>
>
>
> * In ouput file:
> ...
> ## BLOSUM62    : BLOSUM62 substitution score for the reference and
> alternative amino acids     <====
> ## BLOSUM62    : BLOSUM62 substitution score for the reference and
> alternative amino acids     <====
> #Uploaded_variation    Location    Allele    Gene    Feature
> Feature_type    Consequence    cDNA_position
> ...
>
>
>
>
>
>
> Custom example:
> -------------------------
> * In config file:
> custom                xxxxx/vep/ucsc_encode/All_hg19_RS.bw,GERP_RS,bigwig
>
>
> * Shell log:
> [15-07-22 22:43:57]  > ./variant_effect_predictor.pl --config
> cache/vep.ini -i example_GRCh37.vcf
> ...
> version 81
>
> By Will McLaren (wm2 at ebi.ac.uk)
> ...
> custom            xxxxx/vep/ucsc_encode/All_hg19_RS.bw,GERP_RS,bigwig
> xxxxx/vep/ucsc_encode/All_hg19_RS.bw,GERP_RS,bigwig
> ...
>
>   * In ouput file:
> ...
> ## GERP_RS : xxxxx/vep/ucsc_encode/All_hg19_RS.bw (exact)
> ## GERP_RS : xxxxx/vep/ucsc_encode/All_hg19_RS.bw (exact)
>
> ...
>
>
> Best,
> Kirsley
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20150723/257edac4/attachment.html>