XrefParser::SGDParser Class Reference
Public Member Functions | |
| public | run () |
Detailed Description
Definition at line 5 of file SGDParser.pm.
Member Function Documentation
◆ run()
| public XrefParser::SGDParser::run | ( | ) |
Undocumented method
Code:
click to view
sub run {
my ($self, $ref_arg) = @_;
my $source_id = $ref_arg->{source_id};
my $species_id = $ref_arg->{species_id};
my $files = $ref_arg->{files};
my $verbose = $ref_arg->{verbose};
my $dbi = $ref_arg->{dbi};
$dbi = $self->dbi unless defined $dbi;
if((!defined $source_id) or (!defined $species_id) or (!defined $files) ){
croak "Need to pass source_id, species_id and files as pairs";
}
$verbose |=0;
my $file = @{$files}[0];
my $gene_source_id = $self->get_source_id_for_source_name("SGD_GENE", undef, $dbi);
#my $transcript_source_id = $self->get_source_id_for_source_name("SGD_TRANSCRIPT");
my $translation_source_id = $self->get_source_id_for_source_name("SGD_TRANSLATION", undef, $dbi);
my $sgd_io = $self->get_filehandle($file);
if ( !defined $sgd_io ) {
print STDERR "ERROR: Could not open $file\n";
return 1; # 1 is an error
}
my $xref_count =0;
my $syn_count =0;
while ( $_ = $sgd_io->getline() ) {
chomp;
if ($_ =~ /^([^\t]+)\t([^\t]+)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t([^\t]*)\t[^\t]+\t[^\t]*\t[^\t]+\t[^\t]+\t[^\t]+\t[^\t]+\t[^\t]*\t[^\t]+\t[^\t]+\t([^\t]*)$/) {
my ($sgd_id, $biotype, $status, $orf_name, $locus_name, $alias_name, $desc) = ($1,$2,$3,$4,$5,$6,$7);
# parse the lines corresponding to the gene entries
# and filter out lines corresponding to the CDS for example
if ($biotype =~ /ORF|.+RNA|transposable_element_gene|pseudogene/) {
if ($verbose) {
#print STDERR "parsing line for biotype, $biotype\n";
#print STDERR "sgd_id, biotype, status, orf_name, locus_name, alias_name, $sgd_id, $biotype, $status, $orf_name, $locus_name, $alias_name\n";
#print STDERR "desc: $desc\n";
}
if (!defined $locus_name || ($locus_name eq "")) {
if (!defined $orf_name || ($orf_name eq "")) {
print STDERR "can't assign the orf_name as the locus_name!\n";
}
else {
if ($verbose) {
#print STDERR "assigning the orf_name as the locus_name(ie the gene_name)\n";
}
$locus_name = $orf_name;
}
}
my (@syn) = split(/\|/,$alias_name);
my $gene_xref_id = $self->add_xref({ acc => $sgd_id,
label => $locus_name,
desc => $desc,
source_id => $gene_source_id,
species_id => $species_id,
dbi => $dbi,
info_type => "DIRECT"} );
$self->add_direct_xref($gene_xref_id, $orf_name, "Gene", "DIRECT", $dbi);
my $translation_xref_id = $self->add_xref({ acc => $sgd_id,
label => $locus_name,
desc => $desc,
source_id => $translation_source_id,
species_id => $species_id,
dbi => $dbi,
info_type => "DIRECT"} );
$self->add_direct_xref($translation_xref_id, $orf_name, "Translation", "DIRECT", $dbi);
$xref_count++;
foreach my $synonym (@syn){
if ($verbose) {
# print STDERR "adding synonym, $synonym\n";
}
$self->add_to_syn($sgd_id, $gene_source_id, $synonym, $species_id, $dbi);
$syn_count++;
}
}
else {
if ($verbose) {
#print STDERR "filtering biotype, $biotype\n";
}
}
}
else {
if ($verbose) {
print STDERR "failed to parse line, $_\n\n";
}
}
}
$sgd_io->close();
print $xref_count." SGD Xrefs added with $syn_count synonyms\n" if($verbose);
return 0; #successful
}
The documentation for this class was generated from the following file:
- ensembl/misc-scripts/xref_mapping/XrefParser/SGDParser.pm
