Bio/Matrix/PSM/IO/psiblast.pm

   1 #---------------------------------------------------------
   2
   3 =head1 NAME
   4
   5 Bio::Matrix::PSM::IO::psiblast - PSM psiblast parser
   6
   7 =head1 SYNOPSIS
   8
   9 See Bio::Matrix::PSM::IO for documentation
  10
  11 =head1 DESCRIPTION
  12
  13 Parser for ASCII matrices from PSI-BLAST (blastpgp program in
  14 BLAST distribution).
  15
  16 =head1 FEEDBACK
  17
  18 =head2 Mailing Lists
  19
  20 User feedback is an integral part of the evolution of this and other
  21 Bioperl modules. Send your comments and suggestions preferably to one
  22 of the Bioperl mailing lists.  Your participation is much appreciated.
  23
  24   bioperl-l@bioperl.org                  - General discussion
  25   http://bioperl.org/wiki/Mailing_lists  - About the mailing lists
  26
  27 =head2 Support
  28
  29 Please direct usage questions or support issues to the mailing list:
  30
  31 I<bioperl-l@bioperl.org>
  32
  33 rather than to the module maintainer directly. Many experienced and
  34 reponsive experts will be able look at the problem and quickly
  35 address it. Please include a thorough description of the problem
  36 with code and data examples if at all possible.
  37
  38 =head2 Reporting Bugs
  39
  40 Report bugs to the Bioperl bug tracking system to help us keep track
  41 the bugs and their resolution.  Bug reports can be submitted via the
  42 web:
  43
  44   https://github.com/bioperl/bioperl-live/issues
  45
  46 =head1 AUTHOR - James Thompson
  47
  48 Email tex@biosysadmin.com
  49
  50 =head1 APPENDIX
  51
  52 =cut
  53
  54
  55 # Let the code begin...
  56 package Bio::Matrix::PSM::IO::psiblast;
  57 use Bio::Matrix::PSM::Psm;
  58 use Bio::Matrix::PSM::ProtMatrix;
  59 use strict;
  60
  61 use base qw(Bio::Matrix::PSM::PsmHeader Bio::Matrix::PSM::IO);
  62
  63 # define the order in which amino acids are listed in the psiblast matrix file
  64 our @ordered_alphabet = qw/A  R  N  D  C  Q  E  G  H  I  L  K  M  F  P  S  T  W  Y  V/;
  65
  66 =head2 new
  67
  68  Title   : new
  69  Usage   : my $psmIO  =  Bio::Matrix::PSM::IO->new(-format=>'psiblast',
  70                                                   -file=>$file);
  71  Function: Associates a file with the appropriate parser
  72  Throws  :
  73  Example :
  74  Args    :
  75  Returns : Bio::Matrix::PSM::ProtMatrix->new(@args);
  76
  77 =cut
  78
  79 sub new {
  80    my ($class,@args)=@_;
  81    my $line;
  82
  83    my $self   = $class->SUPER::new(@args);
  84    my ($file) = $self->_rearrange(['FILE'], @args);
  85    $self->_initialize_io(@args) || warn "Did you intend to use STDIN?"; # Read only for now
  86    $self->_initialize;
  87
  88    $self->{_ordered_alphabet} = \@ordered_alphabet;
  89    return $self;
  90 }
  91
  92 =head2 next_psm
  93
  94  Title   : next_psm
  95  Usage   : my $psm = $psmIO->next_psm();
  96  Function: Reads the next PSM from the input file, associated with this object
  97  Throws  : None
  98  Returns : Bio::Matrix::PSM::ProtPsm object
  99  Args    : none
 100
 101 =cut
 102
 103 sub next_psm {
 104    my $self = shift;
 105    my $line;
 106
 107    return if ($self->{_end});
 108
 109    my %args;
 110    my @ordered_alphabet = @{$self->{_ordered_alphabet}};
 111
 112    while ( defined( $line = $self->_readline) ) {
 113       # remove leading and trailing whitespace
 114       chomp $line;
 115       $line =~ s/^\s+//g;
 116       $line =~ s/\s+$//g;
 117
 118       if ( $line =~ /^(\d+)\s+(\w{1})/ ) {  # match reference aa and position number
 119          my @elements = split /\s+/, $line;
 120
 121          my $position = shift @elements;
 122          my $letter   = shift @elements;
 123
 124          my $ratio = pop @elements;
 125          my $ic    = pop @elements;
 126
 127          # put the next 20 elements into the correct array in %args
 128          for ( 0 .. 19 ) { push @{$args{'l'.$ordered_alphabet[$_]}}, shift @elements; }
 129          for ( 0 .. 19 ) { push @{$args{'p'.$ordered_alphabet[$_]}}, shift @elements; }
 130
 131          push @{$args{'ic'}}, $ic;
 132       }
 133    }
 134
 135    $self->{_end} = 1; # psiblast matrix files currently only hold one PSM per file
 136
 137    my $psm = Bio::Matrix::PSM::ProtMatrix->new( %args );
 138    return $psm;
 139 }
 140
 141 sub DESTROY {
 142    my $self=shift;
 143    $self->close;
 144 }
 145
 146 1;