branch-1-6/Bio/Tools/RepeatMasker.pm

   1 # $Id$
   2 #
   3 # BioPerl module for Bio::Tools::RepeatMasker
   4 #
   5 # Please direct questions and support issues to <bioperl-l@bioperl.org>
   6 #
   7 # Cared for by Shawn Hoon <shawnh@fugu-sg.org>
   8 #
   9 # Copyright Shawn Hoon
  10 #
  11 # You may distribute this module under the same terms as perl itself
  12
  13 # POD documentation - main docs before the code
  14
  15 =head1 NAME
  16
  17 Bio::Tools::RepeatMasker - a parser for RepeatMasker output
  18
  19 =head1 SYNOPSIS
  20
  21     use Bio::Tools::RepeatMasker;
  22     my $parser = Bio::Tools::RepeatMasker->new(-file => 'seq.fa.out');
  23     while( my $result = $parser->next_result ) {
  24       # get some value
  25     }
  26
  27 =head1 DESCRIPTION
  28
  29 A parser for RepeatMasker output
  30
  31 =head1 FEEDBACK
  32
  33 =head2 Mailing Lists
  34
  35 User feedback is an integral part of the evolution of this and other
  36 Bioperl modules. Send your comments and suggestions preferably to
  37 the Bioperl mailing list.  Your participation is much appreciated.
  38
  39   bioperl-l@bioperl.org                  - General discussion
  40   http://bioperl.org/wiki/Mailing_lists  - About the mailing lists
  41
  42 =head2 Support
  43
  44 Please direct usage questions or support issues to the mailing list:
  45
  46 I<bioperl-l@bioperl.org>
  47
  48 rather than to the module maintainer directly. Many experienced and
  49 reponsive experts will be able look at the problem and quickly
  50 address it. Please include a thorough description of the problem
  51 with code and data examples if at all possible.
  52
  53 =head2 Reporting Bugs
  54
  55 Report bugs to the Bioperl bug tracking system to help us keep track
  56 of the bugs and their resolution. Bug reports can be submitted via
  57 the web:
  58
  59   http://bugzilla.open-bio.org/
  60
  61 =head1 AUTHOR - Shawn Hoon
  62
  63 Email shawnh@fugu-sg.org
  64
  65 =head1 APPENDIX
  66
  67 The rest of the documentation details each of the object methods.
  68 Internal methods are usually preceded with a _
  69
  70 =cut
  71
  72
  73 # Let the code begin...
  74
  75
  76 package Bio::Tools::RepeatMasker;
  77 use strict;
  78
  79 use Bio::SeqFeature::FeaturePair;
  80
  81 use base qw(Bio::Root::Root Bio::Root::IO);
  82
  83 =head2 new
  84
  85  Title   : new
  86  Usage   : my $obj = Bio::Tools::RepeatMasker->new();
  87  Function: Builds a new Bio::Tools::RepeatMasker object
  88  Returns : Bio::Tools::RepeatMasker
  89  Args    : -fh/-file => $val, for initing input, see Bio::Root::IO
  90
  91 =cut
  92
  93 sub new {
  94   my($class,@args) = @_;
  95
  96   my $self = $class->SUPER::new(@args);
  97   $self->_initialize_io(@args);
  98
  99   return $self;
 100 }
 101
 102 =head2 next_result
 103
 104  Title   : next_result
 105  Usage   : my $r = $rpt_masker->next_result
 106  Function: Get the next result set from parser data
 107  Returns : Bio::SeqFeature::FeaturePair
 108            Feature1 is the Query coordinates and Feature2 is the Hit
 109  Args    : none
 110
 111 =cut
 112
 113 sub next_result {
 114     my ($self) = @_;
 115     while (defined($_=$self->_readline()) ) {
 116         if (/no repetitive sequences detected/) {
 117             $self->warn( "RepeatMasker didn't find any repetitive sequences\n");
 118             return ;
 119         }
 120         #ignore introductory lines
 121         if (/\d+/) {
 122             my @element = split;
 123             # ignore features with negatives
 124             next if ($element[11-13] =~ /-/);
 125             my (%feat1, %feat2);
 126             my @line = split;
 127             my ($score, $query_name, $query_start, $query_end, $strand,
 128                 $repeat_name, $repeat_class ) = @line[0, 4, 5, 6, 8, 9, 10];
 129
 130             my ($hit_start,$hit_end);
 131
 132             if ($strand eq '+') {
 133                 ($hit_start, $hit_end) = @line[11, 12];
 134                 $strand = 1;
 135             } elsif ($strand eq 'C') {
 136                 ($hit_start, $hit_end) = @line[12, 13];
 137                 $strand = -1;
 138             }
 139             my $rf = Bio::SeqFeature::Generic->new
 140                 (-seq_id      => $query_name,
 141                  -score       => $score,
 142                  -start       => $query_start,
 143                  -end         => $query_end,
 144                  -strand      => $strand,
 145                  -source_tag  => 'RepeatMasker',
 146                  -primary_tag => $repeat_class,
 147                  -tag => { 'Target'=> [$repeat_name,$hit_start,$hit_end]},
 148                 );
 149
 150             my $rf2 = Bio::SeqFeature::Generic->new
 151                 (-seq_id         => $repeat_name,
 152                  -score          => $score,
 153                  -start          => $hit_start,
 154                  -end            => $hit_end,
 155                  -strand         => $strand,
 156                  -source_tag     => "RepeatMasker",
 157                  -primary_tag    => $repeat_class,
 158                  -tag => { 'Target'=> [$query_name,$query_start,$query_end] },
 159                 );
 160
 161             my $fp = Bio::SeqFeature::FeaturePair->new(-feature1 => $rf,
 162                                                        -feature2 => $rf2);
 163             return $fp;
 164         }
 165     }
 166 }
 167
 168 1;