[ensembl-dev] Exception when retrieving UTR feature from transcript...

Steve Moss gawbul at gmail.com
Mon Apr 15 12:21:53 BST 2013


Dear EnsEMBL Developers,

I receive the following exception when trying to retrieve a UTR feature
from a transcript object. I have run the code on all 61 genomes of release
70 and I can only see issues with 5'-UTRs, but this is probably due to the
data rather than the code and so the 3'-UTR code will likely also be
affected too.

*-------------------- EXCEPTION --------------------
MSG: Must call with start/end
STACK Bio::EnsEMBL::TranscriptMapper::cdna2genomic
/Users/stevemoss/work/ensembl70/ensembl/modules/Bio/EnsEMBL/TranscriptMapper.pm:246
STACK Bio::EnsEMBL::Transcript::cdna2genomic
/Users/stevemoss/work/ensembl70/ensembl/modules/Bio/EnsEMBL/Transcript.pm:2002
STACK Bio::EnsEMBL::Transcript::five_prime_utr_Feature
/Users/stevemoss/work/ensembl70/ensembl/modules/Bio/EnsEMBL/Transcript.pm:1713
STACK toplevel
/Users/stevemoss/Dropbox/PhD/Research/Development/gcat/modules/GCAT/Scripts/
get_intron_counts.pl:124
Date (localtime)    = Mon Apr 15 11:08:05 2013
Ensembl API version = 70
---------------------------------------------------*

The problem seems to be that the five_ and three_prime_utr_Feature method
code assigns *my $cdna_coding = $self->cdna_coding_start();* (or
cda_coding_end for the 3'-UTR) and then assumes a non undef value by then
immediately calling *my ($genomic_pos) = $self->cdna2genomic($cdna_coding,
$cdna_coding);*.

I worked around this by utilising the following code:
*
*
*my $_5putr = defined $tr->cdna_coding_start ? $tr->five_prime_utr_Feature
: undef;
my $_3putr = defined $tr->cdna_coding_end ? $tr->three_prime_utr_Feature :
undef;*

Perhaps this is due to me parsing non-coding genes? I should perhaps just
check the biotype before parsing any further? Any thoughts welcome :)

Kindest regards,

Steve Moss
http://about.me/gawbul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20130415/4b1ec9bf/attachment.html>


More information about the Dev mailing list