Skip to content
Snippets Groups Projects
Commit 13025521 authored by Alessandro Vullo's avatar Alessandro Vullo
Browse files

Merge pull request #87 from james-monkeyshines/master 

Specify 5' or 3' UTR in the GTF file
parents 78c9983c 28154de0
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
modules/Bio/EnsEMBL/Utils/IO/GTFSerializer.pm
+ 2
− 2
View file @ 13025521
......@@ -404,8 +404,8 @@ sub print_feature {
push @$utrs, @{$transcript->get_all_three_prime_UTRs()};
foreach my $utr (@{$utrs}) {
my $strand = $strand_conversion{$utr->strand()};
print $fh sprintf(qq{%s\t%s\tUTR\t%d\t%d\t.\t%s\t.\t},
$idstr, $transcript->source, ($utr->start()+$sliceoffset), ($utr->end+$sliceoffset), $strand);
print $fh sprintf(qq{%s\t%s\t%s\t%d\t%d\t.\t%s\t.\t},
$idstr, $transcript->source, $utr->type, ($utr->start()+$sliceoffset), ($utr->end+$sliceoffset), $strand);
$self->_print_attribs($gene, $biotype_display, $transcript, $transcript_biotype, 0, 'UTR', undef, undef, $has_selenocysteine);
print $fh "\n";
}
......
modules/t/gtfSerialiser.t
+ 4
− 4
View file @ 13025521
......@@ -68,9 +68,9 @@ my $db = $mtdb->get_DBAdaptor("core");
20\tensembl\texon\t30274331\t30274348\t.\t+\t.\tgene_id \"GENE\"; transcript_id \"TRANS\"; exon_number \"1\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; exon_id \"e1\"; tag \"seleno\";
20\tensembl\tCDS\t30274334\t30274345\t.\t+\t0\tgene_id \"GENE\"; transcript_id \"TRANS\"; exon_number \"1\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; protein_id \"PEP\"; tag \"seleno\";
20\tensembl\texon\t30274401\t30274404\t.\t+\t.\tgene_id \"GENE\"; transcript_id \"TRANS\"; exon_number \"2\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; exon_id \"e2\"; tag \"seleno\";
20\tensembl\tUTR\t30274331\t30274333\t.\t+\t.\tgene_id \"GENE\"; transcript_id \"TRANS\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; tag \"seleno\";
20\tensembl\tUTR\t30274346\t30274348\t.\t+\t.\tgene_id \"GENE\"; transcript_id \"TRANS\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; tag \"seleno\";
20\tensembl\tUTR\t30274401\t30274404\t.\t+\t.\tgene_id \"GENE\"; transcript_id \"TRANS\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; tag \"seleno\";
20\tensembl\tfive_prime_utr\t30274331\t30274333\t.\t+\t.\tgene_id \"GENE\"; transcript_id \"TRANS\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; tag \"seleno\";
20\tensembl\tthree_prime_utr\t30274346\t30274348\t.\t+\t.\tgene_id \"GENE\"; transcript_id \"TRANS\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; tag \"seleno\";
20\tensembl\tthree_prime_utr\t30274401\t30274404\t.\t+\t.\tgene_id \"GENE\"; transcript_id \"TRANS\"; gene_source \"ensembl\"; gene_biotype \"protein_coding\"; transcript_source \"ensembl\"; transcript_biotype \"protein_coding\"; tag \"CCDS\"; ccds_id \"CCDS.1\"; tag \"seleno\";
GTF
eq_or_diff(${$fh->string_ref}, $gtf, 'Checking custom Gene object dumps UTRs, Selenocysteine, seleno tag and CCDS');
}
......@@ -135,7 +135,7 @@ my $transcripts_gtf =
20\tvega\tCDS\t30322356\t30322436\t.\t-\t2\tgene_id \"ENSG00000088356\"; gene_version \"1\"; transcript_id \"ENST00000202017\"; transcript_version \"1\"; exon_number \"4\"; gene_name \"C20orf126\"; gene_source \"vega\"; gene_biotype \"protein_coding\"; transcript_name \"C20orf126\"; transcript_source \"vega\"; transcript_biotype \"protein_coding\"; protein_id \"ENSP00000202017\"; protein_version \"1\";
20\tvega\texon\t30320853\t30321749\t.\t-\t.\tgene_id \"ENSG00000088356\"; gene_version \"1\"; transcript_id \"ENST00000202017\"; transcript_version \"1\"; exon_number \"5\"; gene_name \"C20orf126\"; gene_source \"vega\"; gene_biotype \"protein_coding\"; transcript_name \"C20orf126\"; transcript_source \"vega\"; transcript_biotype \"protein_coding\"; exon_id \"ENSE00001109504\"; exon_version \"1\";
20\tvega\tCDS\t30321670\t30321749\t.\t-\t2\tgene_id \"ENSG00000088356\"; gene_version \"1\"; transcript_id \"ENST00000202017\"; transcript_version \"1\"; exon_number \"5\"; gene_name \"C20orf126\"; gene_source \"vega\"; gene_biotype \"protein_coding\"; transcript_name \"C20orf126\"; transcript_source \"vega\"; transcript_biotype \"protein_coding\"; protein_id \"ENSP00000202017\"; protein_version \"1\";
20\tvega\tUTR\t30320853\t30321669\t.\t-\t.\tgene_id \"ENSG00000088356\"; gene_version \"1\"; transcript_id \"ENST00000202017\"; transcript_version \"1\"; gene_name \"C20orf126\"; gene_source \"vega\"; gene_biotype \"protein_coding\"; transcript_name \"C20orf126\"; transcript_source \"vega\"; transcript_biotype \"protein_coding\";
20\tvega\tthree_prime_utr\t30320853\t30321669\t.\t-\t.\tgene_id \"ENSG00000088356\"; gene_version \"1\"; transcript_id \"ENST00000202017\"; transcript_version \"1\"; gene_name \"C20orf126\"; gene_source \"vega\"; gene_biotype \"protein_coding\"; transcript_name \"C20orf126\"; transcript_source \"vega\"; transcript_biotype \"protein_coding\";
",
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment