From 7fe63560a9d6d423abbcad84f8ae7919fe49abb6 Mon Sep 17 00:00:00 2001
From: Glenn Proctor <gp1@sanger.ac.uk>
Date: Mon, 7 Feb 2005 10:59:10 +0000
Subject: [PATCH] Updated to reflect the fact that a CCDS may be linked to more
than one transcript. Hence need to add xrefs only once, i.e. check for
existence before adding.
---
.../xref_mapping/XrefParser/CCDSParser.pm | 24 +++++++++++++++----
1 file changed, 20 insertions(+), 4 deletions(-)
diff --git a/misc-scripts/xref_mapping/XrefParser/CCDSParser.pm b/misc-scripts/xref_mapping/XrefParser/CCDSParser.pm
index ea3a3d4b63..36a09c884f 100644
--- a/misc-scripts/xref_mapping/XrefParser/CCDSParser.pm
+++ b/misc-scripts/xref_mapping/XrefParser/CCDSParser.pm
@@ -2,6 +2,8 @@ package XrefParser::CCDSParser;
use strict;
+use DBI;
+
use XrefParser::BaseParser;
use vars qw(@ISA);
@@ -9,6 +11,8 @@ use vars qw(@ISA);
# Parse file of CCDS records and assign direct xrefs
# All assumed to be linked to transcripts
+# The same CCDS may be linked to more than one transcript, but need to only
+# add the xref once, so check if it already exists before adding it.
sub run {
@@ -16,19 +20,31 @@ sub run {
open(CCDS,"<".$file) || die "Could not open $file\n";
- my $count = 0;
+ my $line_count = 0;
+ my $xref_count = 0;
+
+ my $xref_sth = $self->dbi()->prepare("SELECT xref_id FROM xref WHERE accession=? AND version=? AND source_id=$source_id AND species_id=$species_id");
while (<CCDS>) {
my ($stable_id, $ccds) = split;
- my $xref_id = $self->add_xref($ccds, 1, $ccds, "", $source_id, $species_id);
+ my ($acc, $version) = split (/\./, $ccds);
+ $line_count++;
+
+ # check if an xref already exists
+ $xref_sth->execute($acc, $version);
+ my $xref_id = ($xref_sth->fetchrow_array())[0];
+ if (!$xref_id) {
+ $xref_id = $self->add_xref($acc, $version, $ccds, "", $source_id, $species_id);
+ $xref_count++;
+ }
+
$self->add_direct_xref($xref_id, $stable_id, "transcript", "");
- $count++;
}
- print "Parsed $count CCDS identifiers from $file\n";
+ print "Parsed $line_count CCDS identifiers from $file, added $xref_count xrefs and $line_count direct_xrefs\n";
close(CCDS);
--
GitLab