[ensembl-dev] Issue with V76 Perl API calls now returning GRCh38 instead of GRCh37 coordinates

Giuseppe Gallone gg14 at sanger.ac.uk
Mon May 8 11:28:08 BST 2017


Hi,

We use the Ensembl Perl API to retrieve sets of coordinates for primer 
generation during experimental validation of variants.

Since approximately last week, we are experiencing the following issue 
with our Perl script that retrieves such coordinates: instead of 
returning GRCh37 coordinates as expected, the script is now returning 
GRCh38 coordinates.

We rely on a local copy of the Ensembl API v76, downloaded on 22/8/2014. 
The in-house script that uses this API was written in May 2015.

We are pretty sure nothing has changed in the script or the API, and 
confident the coordinate we were getting so far were GRCh37, because 
these were independently validated prior to primer generation.

Has there been any change introduced which means V.76 API now defaults 
to GRCh38 when invoking a fetch_by_region() call on a slice adaptor?

If so, what would be the best way to return to getting GRCh37 
coordinates from this API version?

For reference, this how the API connection is currently implemented in 
the script:

use Bio::EnsEMBL::Registry;

     sub ensembl_init {
         INFO( "Initialising the EnsEMBL registry to retrieve sequence" );
         Bio::EnsEMBL::Registry->load_registry_from_db(
             -host => 'ensembldb.ensembl.org',
             -user => 'anonymous'
         );

         INFO( "Initialising slice adaptor" );
         $slice_adaptor = Bio::EnsEMBL::Registry->get_adaptor( 'Human', 
'core', 'slice' );

         INFO( "Initialisation complete" );
         return;
     }

     sub get_sequence {
         my ( $chr, $pos, $flank ) = @_;

         $flank ||= 300;

         my $slice = $slice_adaptor->fetch_by_region( 'chromosome', 
$chr, $pos - $flank, $pos + $flank );

         return $slice->seq;
     }


Best wishes,
-- 
Giuseppe Gallone, PhD
Senior Computer Biologist
Genomic mutation and genetic disease
Wellcome Trust Sanger Institute
Wellcome Genome Campus
Hinxton, Cambridge
CB10 1SA, UK



More information about the Dev mailing list