Newer
Older
package XrefParser::ImmunoDBParser;
use strict;
use POSIX qw(strftime);
use File::Basename;
use base qw( XrefParser::BaseParser );
Ian Longden
committed
use Carp;
Ian Longden
committed
my ($self, $ref_arg) = @_;
my $source_id = $ref_arg->{source_id};
my $species_id = $ref_arg->{species_id};
my $files = $ref_arg->{files};
my $release_file = $ref_arg->{rel_file};
my $verbose = $ref_arg->{verbose};
if((!defined $source_id) or (!defined $species_id) or (!defined $files) or (!defined $release_file)){
croak "Need to pass source_id, species_id, files and rel_file as pairs";
}
$verbose |=0;
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
my $file = @{$files}[0];
print "source_id = $source_id, species = $species_id, file = $file\n" if($verbose);
if(!defined($source_id)){
$source_id = XrefParser::BaseParser->get_source_id_for_filename($file);
}
if(!defined($species_id)){
$species_id = XrefParser::BaseParser->get_species_id_for_filename($file);
}
my $added = 0;
my $count = 0;
my $file_io = $self->get_filehandle($file);
if ( !defined $file_io ) {
print STDERR "ERROR: Could not open file $file\n";
return 1;
}
while ( my $line = $file_io->getline() ) {
chomp $line;
my ($SPECIES,$gene_id, $acc, $family, $subfamily, $description) = split(",",$line);
my $full_description = $description."($family)" ;
if ($subfamily ne $family) { $full_description .= ", subfamily $subfamily" ;}
#$subfamily ~= /1-3-beta-D/1,3-beta-D/ ;
my $xref_id = $self->get_xref($acc,$source_id, $species_id);
if(!defined($xref_id)){
$xref_id = $self->add_xref($acc,"", $acc, $full_description, $source_id, $species_id, "DIRECT");
$count++;
}
if(defined($gene_id) and $gene_id ne "-"){
$self->add_direct_xref($xref_id, $gene_id, "Gene", "") ;
$added++;
}
}
$file_io->close();
print "Added $count xrefs and $added Direct xrefs to genes for ImmunoDB\n" if($verbose);
return 0;
}
1;