[ensembl-dev] Differences in Results - Core API vs Website (version 63)
Simon Chan
sichan at bcgsc.ca
Tue Aug 2 20:15:01 BST 2011
I wish to fetch all 3' UTRs from the human genome using the latest version of the core API (version 63), but I'm noticing some differences in the API results and what I see on the website. ENSTs that return nothing from the 'three_prime_utr' method should be treated as non-protein coding RNAs. For ENSG00000187634, the API reports that 3/9 transcripts have 3'UTR (as a result of being protein coding). However, the website reports that 5/9 are protein coding. (See here: http://uswest.ensembl.org/Homo_sapiens/Transcript/ProteinSummary?db=core;g=ENSG00000187634;r=1:860260-879955;t=ENST00000342066 )
My code and results are below. Many thanks for any suggestions!
Simon.
#!/usr/bin/perl -w
use strict;
use Bio::EnsEMBL::Registry;
###
### Access the registry:
###
my $registry = 'Bio::EnsEMBL::Registry';
$registry->load_registry_from_db(
### -host => 'ensembldb.ensembl.org',
### -user => 'anonymous'
-host => 'ensembl01.bcgsc.ca',
-user => 'ensembl',
-pass => 'ensembl',
-port => 3306,
-species => 'homo sapiens',
-verbose => 0,
-db_version => 63
);
my $gene_adaptor = $registry->get_adaptor('Human', 'Core', 'Gene');
my $gene_object = $gene_adaptor->fetch_by_stable_id('ENSG00000187634');
my $all_trans = $gene_object->get_all_Transcripts();
for my $trans_obj ( @{$all_trans} ){
my $ENST = $trans_obj->stable_id;
my $thr_utr = $trans_obj->three_prime_utr;
if (defined $thr_utr){
print "ENST:$ENST\t$thr_utr\n";
} else {
print "ENST:$ENST\tNoUTR\n";
}
}
### Results:
ENST:ENST00000420190 NoUTR
ENST:ENST00000437963 NoUTR
ENST:ENST00000342066 Bio::Seq=HASH(0x182a1850)
ENST:ENST00000341065 Bio::Seq=HASH(0x183fb870)
ENST:ENST00000455979 Bio::Seq=HASH(0x182a0c40)
ENST:ENST00000478729 NoUTR
ENST:ENST00000474461 NoUTR
ENST:ENST00000466827 NoUTR
ENST:ENST00000464948 NoUTR
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20110802/f05352e1/attachment.html>
More information about the Dev
mailing list