Bio/Tools/Eponine.pm

   1 #
   2 # BioPerl module for Bio::Tools::Eponine
   3 #
   4 # Please direct questions and support issues to <bioperl-l@bioperl.org>
   5 #
   6 # Cared for by Tania Oh <gisoht@nus.edu.sg>
   7 #
   8 # Copyright Tania Oh
   9 #
  10 # You may distribute this module under the same terms as perl itself
  11
  12 # POD documentation - main docs before the code
  13
  14 =head1 NAME
  15
  16 Bio::Tools::Eponine - Results of one Eponine run
  17
  18 =head1 SYNOPSIS
  19
  20  use Bio::Tools::Eponine;
  21  use strict;
  22     my $seq = "/data/seq.fa";
  23     my $threshold  = "0.999";
  24     my @params = ( -seq => $seq,
  25                    -threshold => $threshold);
  26
  27    my $factory = Bio::Tools::Run::Eponine->new(@params);
  28      # run eponine against fasta
  29         my $r = $factory->run_eponine($seq);
  30         my $parser = Bio::Tools::Eponine->new($r);
  31
  32        while (my $feat = $parser->next_prediction){
  33                 #$feat contains array of SeqFeature
  34                foreach my $orf($feat) {
  35                    print $orf->seq_id. "\n";
  36                }
  37        }
  38
  39 =head1 DESCRIPTION
  40
  41 Parser for Eponine, a probabilistic transcription start site detector
  42 optimized for mammalian genomic sequence. This module inherits off
  43 Bio::Tools::AnalysisResult and therefore implements
  44 Bio::SeqAnalysisParserI (see L<Bio::Tools::AnalysisResult> and
  45 L<Bio::SeqAnalysisParserI>).
  46
  47 =head1 FEEDBACK
  48
  49 =head2 Mailing Lists
  50
  51 User feedback is an integral part of the evolution of this and other
  52 Bioperl modules. Send your comments and suggestions preferably to one
  53 of the Bioperl mailing lists.  Your participation is much appreciated.
  54
  55   bioperl-l@bioperl.org                  - General discussion
  56   http://bioperl.org/wiki/Mailing_lists  - About the mailing lists
  57
  58 =head2 Support
  59
  60 Please direct usage questions or support issues to the mailing list:
  61
  62 I<bioperl-l@bioperl.org>
  63
  64 rather than to the module maintainer directly. Many experienced and
  65 reponsive experts will be able look at the problem and quickly
  66 address it. Please include a thorough description of the problem
  67 with code and data examples if at all possible.
  68
  69 =head2 Reporting Bugs
  70
  71 Report bugs to the Bioperl bug tracking system to help us keep track
  72 the bugs and their resolution.  Bug reports can be submitted via the
  73 web:
  74
  75   https://github.com/bioperl/bioperl-live/issues
  76
  77 =head1 AUTHOR - Tania Oh
  78
  79 E<lt>gisoht-at-nus.edu.sgE<gt>
  80
  81 =head1 APPENDIX
  82
  83 The rest of the documentation details each of the object methods. Internal methods are usually preceded with a _
  84
  85 =cut
  86
  87
  88 # Let the code begin...
  89
  90
  91 package Bio::Tools::Eponine;
  92 use strict;
  93
  94 use Bio::Tools::Prediction::Gene;
  95 use Bio::Tools::Prediction::Exon;
  96
  97 use base qw(Bio::Tools::AnalysisResult);
  98
  99 sub _initialize_state {
 100     my($self,@args) = @_;
 101
 102     # first call the inherited method!
 103     my $make = $self->SUPER::_initialize_state(@args);
 104
 105     # handle our own parameters
 106
 107     # our private state variables
 108     $self->{'_preds_parsed'} = 0;
 109     #array of Bio::SeqFeatures
 110     $self->{'_flist'} =[];
 111 }
 112
 113 =head2 analysis_method
 114
 115  Usage     : $mzef->analysis_method();
 116  Purpose   : Inherited method. Overridden to ensure that the name matches
 117              /mzef/i.
 118  Returns   : String
 119  Argument  : n/a
 120
 121 =cut
 122
 123 #-------------
 124 sub analysis_method {
 125 #-------------
 126     my ($self, $method) = @_;
 127     if($method && ($method !~ /epo/i)) {
 128         $self->throw("method $method not supported in " . ref($self));
 129     }
 130     return $self->SUPER::analysis_method($method);
 131 }
 132
 133 =head2 next_feature
 134
 135  Title   : next_feature
 136  Usage   : while($gene = $mzef->next_feature()) {
 137                   # do something
 138            }
 139  Function: Returns the next gene structure prediction of the MZEF result
 140            file. Call this method repeatedly until FALSE is returned.
 141
 142            The returned object is actually a SeqFeatureI implementing object.
 143            This method is required for classes implementing the
 144            SeqAnalysisParserI interface, and is merely an alias for
 145            next_prediction() at present.
 146
 147            Note that with the present version of MZEF there will only be one
 148            object returned, because MZEF does not predict individual genes
 149            but just potential internal exons.
 150  Example :
 151  Returns : A Bio::Tools::Prediction::Gene object.
 152  Args    :
 153
 154 =cut
 155
 156 sub next_feature {
 157     my ($self,@args) = @_;
 158     # even though next_prediction doesn't expect any args (and this method
 159     # does neither), we pass on args in order to be prepared if this changes
 160     # ever
 161     return $self->next_prediction(@args);
 162 }
 163
 164 =head2 next_prediction
 165
 166  Title   : next_prediction
 167  Usage   : while($gene = $mzef->next_prediction()) {
 168                   # do something
 169            }
 170  Function: Returns the next gene structure prediction of the MZEF result
 171            file. Call this method repeatedly until FALSE is returned.
 172
 173            Note that with the present version of MZEF there will only be one
 174            object returned, because MZEF does not predict individual genes
 175            but just potential internal exons.
 176  Example :
 177  Returns : A Bio::Tools::Prediction::Gene object.
 178  Args    :
 179
 180 =cut
 181
 182 sub next_prediction {
 183     my ($self) = @_;
 184     my $gene;
 185
 186     # if the prediction section hasn't been parsed yet, we do this now
 187     $self->_parse_predictions() unless $self->_predictions_parsed();
 188
 189     # return the next gene structure (transcript)
 190     return $self->_prediction();
 191 }
 192
 193 =head2 _parse_predictions
 194
 195  Title   : _parse_predictions()
 196  Usage   : $obj->_parse_predictions()
 197  Function: Parses the prediction section. Automatically called by
 198            next_prediction() if not yet done.
 199  Example :
 200  Returns :
 201
 202 =cut
 203
 204 sub _parse_predictions {
 205     my ($self) = @_;
 206
 207     while(defined($_ = $self->_readline())) {
 208         if (! /^\#/){ #ignore introductory lines
 209
 210               my @element = split;
 211               my (%feature);
 212               $feature {name} = $element[0];
 213               $feature {score} = $element[5];
 214               $feature {start} = $element[3];
 215               $feature {end} = $element[4];
 216               $feature {strand} = $element[6];
 217               $feature {source}= 'Eponine';
 218               $feature {primary}= 'TSS';
 219               $feature {program} = 'eponine-scan';
 220               $feature {program_version} = '2';
 221
 222               $self->create_feature(\%feature);
 223                     next;
 224
 225         }
 226     }
 227     $self->_predictions_parsed(1);
 228 }
 229
 230 =head2 create_feature
 231
 232     Title   :   create_feature
 233     Usage   :   obj->create_feature($feature)
 234     Function:   Returns an array of features
 235     Returns :   Returns an array of features
 236     Args    :   none
 237
 238 =cut
 239
 240 sub create_feature {
 241     my ($self, $feat) = @_;
 242      #create and fill Bio::EnsEMBL::Seqfeature object
 243
 244       my $tss = Bio::SeqFeature::Generic->new
 245                     (   -seq_id  => $feat->{'name'},
 246                         -start   => $feat->{'start'},
 247                         -end     => $feat->{'end'},
 248                         -strand  => $feat->{'strand'},
 249                         -score   => $feat->{'score'},
 250                         -source_tag  => $feat->{'source'},
 251                         -primary_tag => $feat->{'primary'});
 252
 253
 254
 255   if ($tss) {
 256          # add to _flist
 257       push(@{$self->{'_flist'}}, $tss);
 258    }
 259
 260    #print $tss->gff_string;
 261 }
 262
 263
 264
 265
 266
 267
 268 =head2 _prediction
 269
 270  Title   : _prediction()
 271  Usage   : $gene = $obj->_prediction()
 272  Function: internal
 273  Example :
 274  Returns :
 275
 276 =cut
 277
 278 sub _prediction {
 279     my ($self) = @_;
 280
 281     return unless(exists($self->{'_flist'}) && @{$self->{'_flist'}});
 282     return shift(@{$self->{'_flist'}});
 283 }
 284
 285 =head2 _predictions_parsed
 286
 287  Title   : _predictions_parsed
 288  Usage   : $obj->_predictions_parsed
 289  Function: internal
 290  Example :
 291  Returns : TRUE or FALSE
 292
 293 =cut
 294
 295 sub _predictions_parsed {
 296     my ($self, $val) = @_;
 297
 298     $self->{'_preds_parsed'} = $val if $val;
 299     # array of pre-parsed predictions
 300     if(! exists($self->{'_preds_parsed'})) {
 301         $self->{'_preds_parsed'} = 0;
 302     }
 303     return $self->{'_preds_parsed'};
 304 }
 305
 306
 307 1;