view release on metacpan or  search on metacpan

lib/Bio/Roary/OrderGenes.pm  view on Meta::CPAN

package Bio::Roary::OrderGenes;
$Bio::Roary::OrderGenes::VERSION = '3.13.0';
# ABSTRACT: Take in GFF files and create a matrix of what genes are beside what other genes


use Moose;
use Bio::Roary::Exceptions;
use Bio::Roary::AnalyseGroups;
use Bio::Roary::ContigsToGeneIDsFromGFF;
use Graph;
use Graph::Writer::Dot;
use File::Basename;

has 'gff_files'                => ( is => 'ro', isa => 'ArrayRef',                  required => 1 );
has 'analyse_groups_obj'       => ( is => 'ro', isa => 'Bio::Roary::AnalyseGroups', required => 1 );
has 'core_definition'          => ( is => 'ro', isa => 'Num',                       default  => 1.0 );
has 'pan_graph_filename'       => ( is => 'ro', isa => 'Str',                       default  => 'core_accessory_graph.dot' );
has 'accessory_graph_filename' => ( is => 'ro', isa => 'Str',                       default  => 'accessory_graph.dot' );
has 'sample_weights'           => ( is => 'ro', isa => 'Maybe[HashRef]' );
has 'samples_to_clusters'      => ( is => 'ro', isa => 'Maybe[HashRef]' );
has 'group_order' => ( is => 'ro', isa => 'HashRef', lazy => 1, builder => '_build_group_order' );
has 'groups_to_sample_names' => ( is => 'rw', isa => 'HashRef', default => sub { {} } );
has 'group_graphs'            => ( is => 'ro', isa => 'Graph',   lazy => 1, builder => '_build_group_graphs' );
has 'groups_to_contigs'       => ( is => 'ro', isa => 'HashRef', lazy => 1, builder => '_build_groups_to_contigs' );
has '_groups_to_file_contigs' => ( is => 'ro', isa => 'HashRef', lazy => 1, builder => '_build__groups_to_file_contigs' );
has '_groups'                 => ( is => 'ro', isa => 'HashRef', lazy => 1, builder => '_build_groups' );
has 'number_of_files'         => ( is => 'ro', isa => 'Int',     lazy => 1, builder => '_build_number_of_files' );
has '_groups_qc' => ( is => 'ro', isa => 'HashRef', default => sub { {} } );
has '_percentage_of_largest_weak_threshold' => ( is => 'ro', isa => 'Num', default => 0.9 );

sub _build_number_of_files {
    my ($self) = @_;
    return @{ $self->gff_files };
}

sub _build_groups {
    my ($self) = @_;
    my %groups;
    for my $group_name ( @{ $self->analyse_groups_obj->_groups } ) {
        $groups{$group_name}++;
    }
    return \%groups;
}

sub _build__groups_to_file_contigs {
    my ($self) = @_;

    my @overlapping_hypothetical_gene_ids;
    my %samples_to_groups_contigs;

    # Open each GFF file
    for my $filename ( @{ $self->gff_files } ) {
        my @groups_to_contigs;
        my $contigs_to_ids_obj = Bio::Roary::ContigsToGeneIDsFromGFF->new( gff_file => $filename );

        my ( $sample_name, $directories, $suffix ) = fileparse($filename);
        $sample_name =~ s/\.gff//gi;

        # Loop over each contig in the GFF file
        for my $contig_name ( keys %{ $contigs_to_ids_obj->contig_to_ids } ) {
            my @groups_on_contig;

            # loop over each gene in each contig in the GFF file
            for my $gene_id ( @{ $contigs_to_ids_obj->contig_to_ids->{$contig_name} } ) {

                # convert to group name
                my $group_name = $self->analyse_groups_obj->_genes_to_groups->{$gene_id};
                next unless ( defined($group_name) );

                if ( $contigs_to_ids_obj->overlapping_hypothetical_protein_ids->{$gene_id} ) {
                    $self->_groups_qc->{$group_name} =
'Hypothetical protein with no hits to refseq/uniprot/clusters/cdd/tigrfams/pfam overlapping another protein with hits';
                }
                push( @groups_on_contig, $group_name );
            }
            push( @groups_to_contigs, \@groups_on_contig );
        }
        $samples_to_groups_contigs{$sample_name} = \@groups_to_contigs;
    }

    return \%samples_to_groups_contigs;

}

sub _build_group_order {
    my ($self) = @_;
    my %group_order;

    my %groups_to_sample_names;