view Roary/lib/Bio/Roary/ExtractCoreGenesFromSpreadsheet.pm @ 0:c47a5f61bc9f draft

Uploaded

package Bio::Roary::ExtractCoreGenesFromSpreadsheet;

# ABSTRACT: Take in a spreadsheet produced by the pipeline and identify the core genes.

=head1 SYNOPSIS

Take in a spreadsheet produced by the pipeline and identify the core genes.
   use Bio::Roary::ExtractCoreGenesFromSpreadsheet;
   
   my $obj = Bio::Roary::ExtractCoreGenesFromSpreadsheet->new(
       spreadsheet        => 'group_statistics.csv',
     );
   $obj->ordered_core_genes();

=cut

use Moose;
use Text::CSV;
use Bio::Roary::GroupStatistics;
use POSIX;

has 'spreadsheet'           => ( is => 'ro', isa  => 'Str',       required => 1 );
has '_csv_parser'           => ( is => 'ro', isa  => 'Text::CSV', lazy     => 1, builder => '_build__csv_parser' );
has '_input_spreadsheet_fh' => ( is => 'ro', lazy => 1,           builder  => '_build__input_spreadsheet_fh' );
has 'ordered_core_genes'    => ( is => 'ro', isa  => 'ArrayRef',  lazy     => 1, builder => '_build_ordered_core_genes' );
has 'core_definition'       => ( is => 'ro', isa  => 'Num',       default  => 1 );
has 'sample_names'          => ( is => 'rw', isa => 'ArrayRef', default => sub { [] } );
has 'sample_names_to_genes' => ( is => 'rw', isa => 'HashRef',  default => sub { {} } );
has 'allow_paralogs'        => ( is => 'rw', isa => 'Bool',     default => 0 );

has '_number_of_isolates'               => ( is => 'rw', isa => 'Int' );
has '_gene_column'                      => ( is => 'rw', isa => 'Int' );
has '_num_isolates_column'              => ( is => 'rw', isa => 'Int' );
has '_avg_sequences_per_isolate_column' => ( is => 'rw', isa => 'Int' );
has '_genome_fragement_column'          => ( is => 'rw', isa => 'Int' );
has '_order_within_fragement_column'    => ( is => 'rw', isa => 'Int' );
has '_min_no_isolates_for_core'         => ( is => 'rw', isa => 'Num', lazy => 1, builder => '_build__min_no_isolates_for_core' );

sub _build__min_no_isolates_for_core {
    my ($self) = @_;
    my $threshold = $self->_number_of_isolates * $self->core_definition;

    return $threshold;
}

sub _build__csv_parser {
    my ($self) = @_;
    return Text::CSV->new( { binary => 1, always_quote => 1 } );
}

sub _build__input_spreadsheet_fh {
    my ($self) = @_;
    open( my $fh, $self->spreadsheet );
    return $fh;
}

sub _update_number_of_isolates {
    my ( $self, $header_row ) = @_;
    my $number_of_isolates = @{$header_row} - @{ Bio::Roary::GroupStatistics->fixed_headers };
    $self->_number_of_isolates($number_of_isolates);
}

sub _setup_column_mappings {
    my ( $self, $header_row ) = @_;

    # current ordering
    my %columns_of_interest_mappings = (
        'Gene'                      => 0,
        'No. isolates'              => 3,
        'Avg sequences per isolate' => 5,
        'Genome Fragment'           => 6,
        'Order within Fragment'     => 7,
        'QC'                        => 10,
    );

    # Dynamically overwrite the default ordering
    for ( my $i = 0 ; $i < @{$header_row} ; $i++ ) {
        for my $col_name (%columns_of_interest_mappings) {
            if ( $header_row->[$i] eq $col_name ) {
                $columns_of_interest_mappings{$col_name} = $i;
                last;
            }
        }
    }
    $self->_gene_column( $columns_of_interest_mappings{'Gene'} );
    $self->_num_isolates_column( $columns_of_interest_mappings{'No. isolates'} );
    $self->_avg_sequences_per_isolate_column( $columns_of_interest_mappings{'Avg sequences per isolate'} );
    $self->_genome_fragement_column( $columns_of_interest_mappings{'Genome Fragment'} );
    $self->_order_within_fragement_column( $columns_of_interest_mappings{'Order within Fragment'} );
    $self->_update_number_of_isolates($header_row);

    # Get the sample_names
    my @sample_names;
    for ( my $i = $self->_length_of_fixed_headers() ; $i < @{$header_row} ; $i++ ) {
        push( @sample_names, $header_row->[$i] );
    }
    $self->sample_names( \@sample_names );
}

sub _length_of_fixed_headers {
    my ($self) = @_;
    return @{ Bio::Roary::GroupStatistics->fixed_headers() };
}

sub _populate_sample_to_gene_lookup_with_row {
    my ( $self, $row ) = @_;

    for ( my $i = $self->_length_of_fixed_headers() ; $i < @{$row} ; $i++ ) {
        if ( defined( $row->[$i] ) && $row->[$i] ne "" ) {
            my $sample_name = $self->sample_names->[ $i - $self->_length_of_fixed_headers() ];

            $self->sample_names_to_genes->{$sample_name}->{ $row->[$i] } = 1;
        }
    }
    return 1;
}

sub _ordered_core_genes {
    my ($self) = @_;
    my %ordered_genes;
    while ( my $row = $self->_csv_parser->getline( $self->_input_spreadsheet_fh ) ) {
        next if ( @{$row} < 12 );                                                                        # no genes in group
        next if ( !defined( $row->[ $self->_gene_column ] ) || $row->[ $self->_gene_column ] eq '' );    # no gene name
        next
          if ( !defined( $row->[ $self->_avg_sequences_per_isolate_column ] ) || $row->[ $self->_avg_sequences_per_isolate_column ] eq '' )
          ;                                                                                              # no average
        next
          if ( !defined( $row->[ $self->_genome_fragement_column ] ) || $row->[ $self->_genome_fragement_column ] eq '' )
          ;                                                                                              # fragment not defined

        # next if($self->_number_of_isolates != $row->[$self->_num_isolates_column]); # if gene is not in all isolates
        next if ( $row->[ $self->_num_isolates_column ] < $self->_min_no_isolates_for_core );

        if ( $self->allow_paralogs ) {
            # should never happen
            next if ( $row->[ $self->_avg_sequences_per_isolate_column ] < 1 );
        }
        else {
            next if ( $row->[ $self->_avg_sequences_per_isolate_column ] != 1 );
        }

        $ordered_genes{ $row->[ $self->_genome_fragement_column ] }{ $row->[ $self->_order_within_fragement_column ] } =
          $row->[ $self->_gene_column ];
        $self->_populate_sample_to_gene_lookup_with_row($row);
    }

    my @ordered_core_genes;
    for my $fragment_key ( sort { $a <=> $b } keys %ordered_genes ) {
        for my $order_within_fragement ( sort { $a <=> $b } keys %{ $ordered_genes{$fragment_key} } ) {
            push( @ordered_core_genes, $ordered_genes{$fragment_key}{$order_within_fragement} );
        }
    }
    return \@ordered_core_genes;
}

sub _build_ordered_core_genes {
    my ($self) = @_;
    my $header_row = $self->_csv_parser->getline( $self->_input_spreadsheet_fh );
    $self->_setup_column_mappings($header_row);

    return $self->_ordered_core_genes();
}

no Moose;
__PACKAGE__->meta->make_immutable;

1;