2 # BioPerl module for Bio::Tools::Sim4::Results
4 # Please direct questions and support issues to <bioperl-l@bioperl.org>
6 # Cared for by Ewan Birney <birney-at-sanger.ac.uk>
7 # and Hilmar Lapp <hlapp-at-gmx.net>
9 # Copyright Ewan Birney and Hilmar Lapp
11 # You may distribute this module under the same terms as perl itself
13 # POD documentation - main docs before the code
17 Bio::Tools::Sim4::Results - Results of one Sim4 run
21 # to preset the order of EST and genomic file as given on the sim4
23 my $sim4 = Bio::Tools::Sim4::Results->new(-file => 'result.sim4',
25 # to let the order be determined automatically (by length comparison):
26 $sim4 = Bio::Tools::Sim4::Results->new( -file => 'sim4.results' );
28 $sim4 = Bio::Tools::Sim4::Results->new( -fh => \*INPUT );
31 while(my $exonset = $sim4->next_exonset()) {
32 # $exonset is-a Bio::SeqFeature::Generic with Bio::Tools::Sim4::Exons
34 print "Delimited on sequence ", $exonset->seq_id(),
35 "from ", $exonset->start(), " to ", $exonset->end(), "\n";
36 foreach my $exon ( $exonset->sub_SeqFeature() ) {
37 # $exon is-a Bio::SeqFeature::FeaturePair
38 print "Exon from ", $exon->start, " to ", $exon->end,
39 " on strand ", $exon->strand(), "\n";
40 # you can get out what it matched using the est_hit attribute
41 my $homol = $exon->est_hit();
42 print "Matched to sequence ", $homol->seq_id,
43 " at ", $homol->start," to ", $homol->end, "\n";
47 # essential if you gave a filename at initialization (otherwise the file
53 The sim4 module provides a parser and results object for sim4 output. The
54 sim4 results are specialised types of SeqFeatures, meaning you can add them
55 to AnnSeq objects fine, and manipulate them in the "normal" seqfeature manner.
57 The sim4 Exon objects are Bio::SeqFeature::FeaturePair inherited objects. The
58 $esthit = $exon-E<gt>est_hit() is the alignment as a feature on the matching
59 object (normally, an EST), in which the start/end points are where the hit
62 To make this module work sensibly you need to run
64 sim4 genomic.fasta est.database.fasta
66 sim4 est.fasta genomic.database.fasta
68 To get the sequence identifiers recorded for the first sequence, too, use
69 A=4 as output option for sim4.
71 One fiddle here is that there are only two real possibilities to the matching
72 criteria: either one sequence needs reversing or not. Because of this, it
73 is impossible to tell whether the match is in the forward or reverse strand
74 of the genomic DNA. We solve this here by assuming that the genomic DNA is
75 always forward. As a consequence, the strand attribute of the matching EST is
76 unknown, and the strand attribute of the genomic DNA (i.e., the Exon object)
77 will reflect the direction of the hit.
79 See the documentation of parse_next_alignment() for abilities of the parser
80 to deal with the different output format options of sim4.
86 User feedback is an integral part of the evolution of this and other
87 Bioperl modules. Send your comments and suggestions preferably to one
88 of the Bioperl mailing lists. Your participation is much appreciated.
90 bioperl-l@bioperl.org - General discussion
91 http://bioperl.org/wiki/Mailing_lists - About the mailing lists
95 Please direct usage questions or support issues to the mailing list:
97 I<bioperl-l@bioperl.org>
99 rather than to the module maintainer directly. Many experienced and
100 reponsive experts will be able look at the problem and quickly
101 address it. Please include a thorough description of the problem
102 with code and data examples if at all possible.
104 =head2 Reporting Bugs
106 Report bugs to the Bioperl bug tracking system to help us keep track
107 the bugs and their resolution. Bug reports can be submitted via the
110 https://github.com/bioperl/bioperl-live/issues
112 =head1 AUTHOR - Ewan Birney, Hilmar Lapp
114 Ewan Birney E<lt>birney-at-sanger.ac.ukE<gt>
115 Hilmar Lapp E<lt>hlapp-at-gmx.netE<gt> or E<lt>hilmar.lapp-at-pharma.novartis.comE<gt>.
119 The rest of the documentation details each of the object methods. Internal methods are usually preceded with a _
124 # Let the code begin...
127 package Bio
::Tools
::Sim4
::Results
;
133 use Bio
::Tools
::Sim4
::Exon
;
135 use base
qw(Bio::Tools::AnalysisResult);
138 sub _initialize_state
{
139 my($self,@args) = @_;
141 # call the inherited method first
142 my $make = $self->SUPER::_initialize_state
(@args);
144 my ($est_is_first) = $self->_rearrange([qw(ESTFIRST)], @args);
146 delete($self->{'_est_is_first'});
147 $self->{'_est_is_first'} = $est_is_first if(defined($est_is_first));
148 $self->analysis_method("Sim4");
151 =head2 analysis_method
153 Usage : $sim4->analysis_method();
154 Purpose : Inherited method. Overridden to ensure that the name matches
162 sub analysis_method
{
164 my ($self, $method) = @_;
165 if($method && ($method !~ /sim4/i)) {
166 $self->throw("method $method not supported in " . ref($self));
168 return $self->SUPER::analysis_method
($method);
171 =head2 parse_next_alignment
173 Title : parse_next_alignment
174 Usage : @exons = $sim4_result->parse_next_alignment;
175 foreach $exon (@exons) {
178 Function: Parses the next alignment of the Sim4 result file and returns the
179 found exons as an array of Bio::Tools::Sim4::Exon objects. Call
180 this method repeatedly until an empty array is returned to get the
181 results for all alignments.
183 The $exon->seq_id() attribute will be set to the identifier of the
184 respective sequence for both sequences if A=4 was used in the sim4
185 run, and otherwise for the second sequence only. If the output does
186 not contain the identifier, the filename stripped of path and
187 extension is used instead. In addition, the full filename
188 will be recorded for both features ($exon inherits off
189 Bio::SeqFeature::SimilarityPair) as tag 'filename'. The length
190 is accessible via the seqlength() attribute of $exon->query() and
193 Note that this method is capable of dealing with outputs generated
194 with format 0,1,3, and 4 (via the A=n option to sim4). It
195 automatically determines which of the two sequences has been
196 reversed, and adjusts the coordinates for that sequence. It will
197 also detect whether the EST sequence(s) were given as first or as
198 second file to sim4, unless this has been specified at creation
202 Returns : An array of Bio::Tools::Sim4::Exon objects
208 sub parse_next_alignment
{
213 # we refer to the properties of each seq by reference
214 my ($estseq, $genomseq, $to_reverse);
216 my $hit_direction = 1;
217 my $output_fmt = 3; # same as 0 and 1 (we cannot deal with A=2 produced
220 while(defined($_ = $self->_readline())) {
223 # bascially, each sim4 'hit' starts with seq1...
227 $self->_pushback($_);
232 # filename and length of seq 1
233 /^seq1\s+=\s+(\S+)\,\s+(\d+)/ ||
234 $self->throw("Sim4 parsing error on seq1 [$_] line. Sorry!");
235 $seq1props{'filename'} = $1;
236 $seq1props{'length'} = $2;
240 # the second hit has also the database name in the >name syntax
242 /^seq2\s+=\s+(\S+)\s+\(>?(\S+)\s*\)\,\s+(\d+)/||
243 $self->throw("Sim4 parsing error on seq2 [$_] line. Sorry!");
244 $seq2props{'filename'} = $1;
245 $seq2props{'seqname'} = $2;
246 $seq2props{'length'} = $3;
249 if(/^>(\S+)\s*(.*)$/) {
250 # output option was A=4, which not only gives the complete
251 # description lines, but also causes the longer sequence to be
252 # reversed if the second file contained one (genomic) sequence
253 $seq1props{'seqname'} = $1;
254 $seq1props{'description'} = $2 if $2;
256 # we handle seq1 and seq2 both here
257 if(defined($_ = $self->_readline()) && (/^>(\S+)\s*(.*)$/)) {
258 $seq2props{'seqname'} = $1; # redundant, since already set above
259 $seq2props{'description'} = $2 if $2;
263 /^\(complement\)/ && do {
268 # start-end (start-end) pctid%
269 if(/(\d+)-(\d+)\s+\((\d+)-(\d+)\)\s+(\d+)%/) {
270 $seq1props{'start'} = $1;
271 $seq1props{'end'} = $2;
272 $seq2props{'start'} = $3;
273 $seq2props{'end'} = $4;
276 if(! defined($estseq)) {
277 # for the first time here: need to set the references referring
279 if(! exists($self->{'_est_is_first'})) {
280 # detect which one is the EST by looking at the lengths,
281 # and assume that this holds throughout the entire result
282 # file (i.e., when this method is called for the next
283 # alignment, this will not be checked again)
284 if($seq1props{'length'} > $seq2props{'length'}) {
285 $self->{'_est_is_first'} = 0;
287 $self->{'_est_is_first'} = 1;
290 if($self->{'_est_is_first'}) {
291 $estseq = \
%seq1props;
292 $genomseq = \
%seq2props;
293 # if the EST is given first, A=4 selects the genomic
294 # seq for being reversed (reversing the EST is default)
295 $to_reverse = ($output_fmt == 4) ?
$genomseq : $estseq;
297 $estseq = \
%seq2props;
298 $genomseq = \
%seq1props;
299 # if the EST is the second, A=4 does not change the
300 # seq being reversed (always the EST is reversed)
301 $to_reverse = $estseq;
304 if($hit_direction == -1) {
305 # we have to reverse the coordinates of one of both seqs
306 my $tmp = $to_reverse->{'start'};
307 $to_reverse->{'start'} =
308 $to_reverse->{'length'} - $to_reverse->{'end'} + 1;
309 $to_reverse->{'end'} = $to_reverse->{'length'} - $tmp + 1;
311 # create and initialize the exon object
312 my $exon = Bio
::Tools
::Sim4
::Exon
->new(
313 '-start' => $genomseq->{'start'},
314 '-end' => $genomseq->{'end'},
315 '-strand' => $hit_direction);
316 if(exists($genomseq->{'seqname'})) {
317 $exon->seq_id($genomseq->{'seqname'});
319 # take filename stripped of path as fall back
320 my ($basename) = &File
::Basename
::fileparse
($genomseq->{'filename'}, '\..*');
321 $exon->seq_id($basename);
323 $exon->feature1()->add_tag_value('filename',
324 $genomseq->{'filename'});
325 # feature1 is supposed to be initialized to a Similarity object,
326 # but we provide a safety net
327 if($exon->feature1()->can('seqlength')) {
328 $exon->feature1()->seqlength($genomseq->{'length'});
330 $exon->feature1()->add_tag_value('SeqLength',
331 $genomseq->{'length'});
333 # create and initialize the feature wrapping the 'hit' (the EST)
334 my $fea2 = Bio
::SeqFeature
::Similarity
->new(
335 '-start' => $estseq->{'start'},
336 '-end' => $estseq->{'end'},
338 '-primary' => "aligning_EST");
339 if(exists($estseq->{'seqname'})) {
340 $fea2->seq_id($estseq->{'seqname'});
342 # take filename stripped of path as fall back
344 &File
::Basename
::fileparse
($estseq->{'filename'}, '\..*');
345 $fea2->seq_id($basename);
347 $fea2->add_tag_value('filename', $estseq->{'filename'});
348 $fea2->seqlength($estseq->{'length'});
350 $exon->est_hit($fea2);
352 $exon->source_tag($self->analysis_method());
353 $exon->percentage_id($pctid);
354 $exon->score($exon->percentage_id());
357 next; # back to while loop
366 Usage : $exonset = $sim4_result->parse_next_exonset;
367 print "Exons start at ", $exonset->start(),
368 "and end at ", $exonset->end(), "\n";
369 foreach $exon ($exonset->sub_SeqFeature()) {
372 Function: Parses the next alignment of the Sim4 result file and returns the
373 set of exons as a container of features. The container is itself
374 a Bio::SeqFeature::Generic object, with the Bio::Tools::Sim4::Exon
375 objects as sub features. Start, end, and strand of the container
376 will represent the total region covered by the exons of this set.
378 See the documentation of parse_next_alignment() for further
379 reference about parsing and how the information is stored.
382 Returns : An Bio::SeqFeature::Generic object holding Bio::Tools::Sim4::Exon
383 objects as sub features.
392 # get the next array of exons
393 my @exons = $self->parse_next_alignment();
395 return if eof($self->_fh);
396 return $self->next_exonset;
398 # create the container of exons as a feature object itself, with the
399 # data of the first exon for initialization
400 $exonset = Bio
::SeqFeature
::Generic
->new('-start' => $exons[0]->start(),
401 '-end' => $exons[0]->end(),
402 '-strand' => $exons[0]->strand(),
403 '-primary' => "ExonSet");
404 $exonset->source_tag($exons[0]->source_tag());
405 $exonset->seq_id($exons[0]->seq_id());
406 # now add all exons as sub features, with enabling EXPANsion of the region
408 foreach my $exon (@exons) {
409 $exonset->add_sub_SeqFeature($exon, 'EXPAND');
417 Usage : while($exonset = $sim4->next_feature()) {
420 Function: Does the same as L<next_exonset()>. See there for documentation of
421 the functionality. Call this method repeatedly until FALSE is
424 The returned object is actually a SeqFeatureI implementing object.
425 This method is required for classes implementing the
426 SeqAnalysisParserI interface, and is merely an alias for
427 next_exonset() at present.
430 Returns : A Bio::SeqFeature::Generic object.
436 my ($self,@args) = @_;
437 # even though next_exonset doesn't expect any args (and this method
438 # does neither), we pass on args in order to be prepared if this changes
440 return $self->next_exonset(@args);