Var_phenotype.pm

Var_phenotype.pm

=head1 LICENSE
                                                                                                        
 This software is distributed under a modified Apache license.                                                                                                                      
                                                                                                                     
=head1 CONTACT                                                                                                       

 Guillermo Marco
    
=cut

=head1 NAME

=head1 SYNOPSIS

=head1 DESCRIPTION

=cut

package Var_phenotype;

use strict;
use warnings;
use Data::Dumper;
use Bio::EnsEMBL::Variation::Utils::BaseVepPlugin;
use base qw(Bio::EnsEMBL::Variation::Utils::BaseVepPlugin);

sub new {
    my $class = shift;    
    my $self = $class->SUPER::new(@_);
    $self->{config}->{check_existing} = 1;
    
    return $self;
}

sub version {
    return '1.0';
}

sub feature_types {
    return ['Feature'];
}

sub get_header_info {
    return {
        Phenotype_association => "Phentoype asociation description",
        ClinVar_ID => "ClinVar ID",
        ClinVar_Sig_Level => "ClinVar significance level",
        ClinVar_Clinical_Sig => "ClinVar clinical significance"
    };
}

sub parse_description{
    my $string = shift;
    $string =~ s/^\s+|\s+$//g;
    $string =~ s/,/%25/g;
    $string =~ s/ /%20/g;
    $string =~ s/;/%27/g;
    $string =~ s/\|/%21/g;
    $string =~ s/&/%26/g;
    $string =~ s/@/%22/g;
    $string =~ s/_+/_/g;
    $string =~ s/[^a-zA-Z0-9_%-]//g;
    return $string;
}

sub run {
    $Data::Dumper::Maxdepth = 3;
    ## Def vars
    my %res;
    my @phenotype_descriptions = ();
    my @clinvar_ids = ();
    my @clinvar_is_significant = ();
    my @clinvar_clinical_sig = ();

    my ($self, $tva) = @_;
    ## Def Handlers
    my $vf = $tva->variation_feature;
    my $pfa = $self->{config}->{reg}->get_adaptor('human','variation','phenotypefeature');

    my $omim_counter = 0;
    my $clinvar_counter = 0;
    foreach my $known_var(@{$vf->{existing} || []}) {
        foreach my $pf(@{$pfa->fetch_all_by_object_id($known_var->{variation_name})}){
            @phenotype_descriptions = ();
            if ($pf->source_name =~ /clinvar/i && defined $pf->risk_allele){
                if ($pf->risk_allele eq $tva->variation_feature_seq){
                    push(@clinvar_ids, $pf->external_id) if (defined($pf->external_id));
                    push(@clinvar_is_significant, $pf->is_significant) if (defined($pf->is_significant));
                    push(@clinvar_clinical_sig, parse_description($pf->clinical_significance)) if (defined($pf->clinical_significance));
                }
            }
            elsif ($pf->source_name =~ /omim/i){
                if($pf->phenotype->description !~ @phenotype_descriptions){
                    #Todo: atm Ensembl has no MIM ID for variant-phenotype association
                    #Todo: when they implement it we should distinguish between OMIM and Clinvar descrptions and IDs
                    push(@phenotype_descriptions, parse_description($pf->phenotype->description))
                }
            }
        }
    }
    $res{"Phenotype_association"} = join('&', @phenotype_descriptions);
    $res{"ClinVar_ID"} = join('&', @clinvar_ids);
    $res{"ClinVar_Sig_Level"} = join('&', @clinvar_is_significant);
    $res{"ClinVar_Clinical_Sig"} = join('&', @clinvar_clinical_sig);
    return { %res };
}
1;