diff --git a/modules/Bio/EnsEMBL/EvidenceAlignment.pm b/modules/Bio/EnsEMBL/EvidenceAlignment.pm
deleted file mode 100755
index 502bca072ae26f636115c245d050bdc0b82868c6..0000000000000000000000000000000000000000
--- a/modules/Bio/EnsEMBL/EvidenceAlignment.pm
+++ /dev/null
@@ -1,1464 +0,0 @@
-#
-#
-# Cared for by Ensembl <ensembl-dev@ebi.ac.uk>
-#
-# You may distribute the source code of the module
-# under the same terms as perl itself.
-#
-# POD documentation - main docs before the code
-
-=pod
-
-=head1 NAME
-
-Bio::EnsEMBL::EvidenceAlignment.pm
-
-=head1 SYNOPSIS
-
- my $ea = Bio::EnsEMBL::EvidenceAlignment->new(
- -DBADAPTOR => $dba,
- -TRANSCRIPTID => $tr_stable_id,
- -PFETCH => '/path/to/pfetch');
- my @alignment_arr = $ea->fetch_alignment;
- $ea->transcriptid($other_tr_dbID);
- my @other_alignment_arr = $ea->fetch_alignment;
- my @short_alignment_arr = $ea->fetch_alignment($hid);
- $ea->contigname($contig_name);
- my @contig_based_alignment_arr = $ea->fetch_alignment;
-
-=head1 DESCRIPTION
-
-Gives an alignment of either a transcript and its evidence or a raw
-contig and its similarity features. Coordinates in the database are
-used to recreate the alignment, and must be correct or some data may
-not be displayed or may be displayed incorrectly.
-
-=head1 CONTACT
-
-Ensembl: ensembl-dev@ebi.ac.uk
-
-=head1 APPENDIX
-
-The rest of the documentation details each of the object
-methods. Internal methods are usually preceded with a _
-
-=cut
-
-package Bio::EnsEMBL::EvidenceAlignment;
-
-# modify slices' features' genomic start/end by the following:
-use constant VC_HACK_BP => +0;
-
-use vars qw(@ISA);
-use strict;
-use Bio::EnsEMBL::DBSQL::DBAdaptor;
-use Bio::SeqIO;
-use Bio::Seq;
-use Bio::EnsEMBL::Gene;
-use Bio::EnsEMBL::Root;
-
-@ISA = qw(Bio::EnsEMBL::Root);
-
-=head2 new
-
- Title : new
- Usage : my $tr_ea = Bio::EnsEMBL::EvidenceAlignment->new(
- -DBADAPTOR => $dba,
- -TRANSCRIPTID => $transcript_id,
- -SEQFETCHER => $seqfetcher,
- -USE_SUPPORTING_EVIDENCE => 1);
- my $cont_ea = Bio::EnsEMBL::EvidenceAlignment->new(
- -DBADAPTOR => $dba,
- -CONTIGID => $contig_stable_id);
- Function: Initialises EvidenceAlignment object
- Returns : An EvidenceAlignment object
- Args : Database adaptor object and an ID string (-CONTIGID
- with contig ID or -TRANSCRIPTID with transcript
- ID, which may be either the transcript stable ID or
- the transcript dbID); optional full path of pfetch
- executable (-PFETCH), defaulting to whatever is in the
- users search path; USE_SUPPORTING_EVIDENCE which, if
- true, causes retrieval of supporting evidence rather
- than similarity features for transcripts (and has no
- effect for contigs), defaulting to undef; optional
- SeqFetcher (-SEQFETCHER), which, if present, will be
- used instead of the pfetch executable. The seqfetcher,
- if supplied, must have a get_Seqs_by_accs method.
-
-=cut
-
-sub new {
- my($class,@args) = @_;
-
- my $self = $class->SUPER::new(@args);
- my ($transcriptid, $contigname, $dbadaptor, $seqfetcher, $pfetch,
- $use_supporting_evidence) =
- $self->_rearrange(['TRANSCRIPTID',
- 'CONTIGID',
- 'DBADAPTOR',
- 'SEQFETCHER',
- 'PFETCH',
- 'USE_SUPPORTING_EVIDENCE'],
- @args);
-
- if (defined $transcriptid and defined $contigname) {
- $self->throw("may have a transcript ID or a contig name but not both");
- }
- if ($transcriptid) {
- $self->transcriptid($transcriptid);
- }
- if ($contigname) {
- $self->contigname($contigname);
- }
- if (! $pfetch) {
- $pfetch = 'pfetch';
- }
- $self->pfetch($pfetch);
- $self->use_supporting_evidence($use_supporting_evidence);
- $self->dbadaptor($dbadaptor);
-
- # If seqfetcher not specified, we leave it undefined, and pfetch
- # executable will be used instead.
- $self->seqfetcher($seqfetcher);
-
- return $self; # success - we hope!
-}
-
-=head2 dbadaptor
-
- Title : dbadaptor
- Usage : $ea->dbadaptor($dba);
- Function: get/set for database adaptor object
-
-=cut
-
-sub dbadaptor {
- my $obj = shift;
- if( @_ ) {
- my $value = shift;
- $obj->{evidencealignment_db_adaptor} = $value;
- }
- return $obj->{evidencealignment_db_adaptor};
-}
-
-=head2 seqfetcher
-
- Title : seqfetcher
- Usage : $ea->seqfetcher($seqfetcher_obj);
- Function: get/set for seqfetcher, which must have a
- get_Seqs_by_accs method
-
-=cut
-
-sub seqfetcher {
- my $obj = shift;
- if( @_ ) {
- my $value = shift;
- $obj->{evidencealignment_seqfetcher} = $value;
- }
- return $obj->{evidencealignment_seqfetcher};
-}
-
-=head2 pfetch
-
- Title : pfetch
- Usage : $ea->pfetch('/path/to/pfetch');
- Function: get/set for pfetch, the location of the pfetch
- binary.
-
-=cut
-
-sub pfetch {
- my $obj = shift;
- if( @_ ) {
- my $value = shift;
- $obj->{evidencealignment_pfetch} = $value;
- }
- return $obj->{evidencealignment_pfetch};
-}
-
-=head2 transcriptid
-
- Title : transcriptid
- Usage : $ea->transcriptid($transcript_stable_id);
- Function: get/set for transcript id string, which may be
- either the stable ID or the internal (db) ID
- (setting transcriptid also unsets contigid)
-
-=cut
-
-sub transcriptid {
- my $obj = shift;
- if( @_ ) {
- my $value = shift;
- $obj->{evidencealignment_transcript_id} = $value;
- undef $obj->{evidencealignment_contig_name};
- }
- return $obj->{evidencealignment_transcript_id};
-}
-
-=head2 use_supporting_evidence
-
- Title : use_supporting_evidence
- Usage : $ea->use_supporting_evidence(1);
- Function: Get/set for use_supporting_evidence. True
- means that, for transcripts, we obtain
- supporting evidence for each exon.
- Otherwise, we obtain all similarity features
- that overlap any exon on the same strand.
- For contigs, this option has no effect.
-
-=cut
-
-sub use_supporting_evidence {
- my $obj = shift;
- if( @_ ) {
- my $value = shift;
- $obj->{evidencealignment_use_supporting_evidence} = $value;
- }
- return $obj->{evidencealignment_use_supporting_evidence};
-}
-
-=head2 contigname
-
- Title : contigname
- Usage : $ea->contigname($contig_name);
- Function: get/set for contig name string (setting this
- also unsets transcriptid)
-
-=cut
-
-sub contigname {
- my $obj = shift;
- if( @_ ) {
- my $value = shift;
- $obj->{evidencealignment_contig_name} = $value;
- undef $obj->{evidencealignment_transcript_id};
- }
- return $obj->{evidencealignment_contig_name};
-}
-
-=head2 _get_evidence_from_transcript
-
- Title : _get_evidence_from_transcript
- Usage : $ea->_get_evidence_from_transcript($transcript_obj);
- $ea->_get_evidence_from_transcript($transcript_obj,
- $hid);
- Function: to get "supporting evidence" (actually similarity
- features) for the transcript supplied, by taking
- them off a slice; features not overlapping any exon
- are cut; if a list of hit accession numbers is
- given, features not involving those accession
- numbers are cut; duplicate features are removed
- Returns : array of featurepairs
-
-=cut
-
-sub _get_evidence_from_transcript {
- my ($self, $transcript_obj) = splice @_, 0, 2;
- $self->throw('interface fault') if (!$self or !$transcript_obj);
- my @wanted_arr = @_;
-
- my @exons = @{$transcript_obj->get_all_Exons};
- my $strand = $exons[0]->strand;
- my @features = ();
- foreach my $exon (@exons) {
- my $exon_features = $exon->get_all_supporting_features;
- foreach my $feature (@$exon_features) {
- next unless $feature->primary_tag =~ /similarity/i;
- # store on overlap, unless we're not interested in this hit seq.
- if ($exon->overlaps($feature)) { # which it should (evidence)!
- my $hid = $feature->hseqname;
- my $wanted = 0;
- if (@wanted_arr) {
- foreach (@wanted_arr) { # we only want some hits
- if ($_ eq $hid) {
- $wanted = 1;
- }
- }
- } else { # we want all hits
- $wanted = 1;
- }
- if ($wanted) {
- push @features, $feature;
- }
- }
- }
- }
-
- my $nonredundant_f_arr_ref = $self->_remove_duplicate_features(
- \@features);
- return @$nonredundant_f_arr_ref;
-}
-
-=head2 _get_similarity_features_from_transcript
-
- Title : _get_similarity_features_from_transcript
- Usage : $ea->_get_similarity_features_from_transcript(
- $transcript_obj, $slice);
- $ea->_get_similarity_features_from_transcript(
- $transcript_obj, $slice, $hid);
- Function: use SGP adaptor supplied to get similarity features
- off a slice of the transcript supplied; features not
- overlapping any exon are cut; if a list of hit
- accession numbers are given, features not involving
- those accession numbers are cut; genomic start and
- end are always modified by VC_HACK_BP; duplicate
- features are removed
- Returns : array of featurepairs
-
-=cut
-
-sub _get_similarity_features_from_transcript {
- my ($self, $transcript_obj, $slice) = splice @_, 0, 3;
- $self->throw('interface fault')
- if (!$self or !$transcript_obj or !$slice);
- my @wanted_arr = @_;
-
- my @exons = @{$transcript_obj->get_all_Exons};
- my $strand = $exons[0]->strand;
- my $db = $self->dbadaptor;
- my $pfadp = $db->get_ProteinAlignFeatureAdaptor;
- my @gapped_features = @{$pfadp->fetch_all_by_Slice($slice)};
- my $dfadp = $db->get_DnaAlignFeatureAdaptor;
- push @gapped_features, @{$dfadp->fetch_by_Slice($slice)};
- my @all_features = ();
- foreach my $gapped_feature (@gapped_features) {
- push @all_features, $gapped_feature->_parse_cigar;
- }
-
- my @features = ();
- FEATURE_LOOP:
- foreach my $feature (@all_features) {
- next unless $feature->primary_tag =~ /similarity/i;
- if ($feature->strand == $strand) {
- # fix VC-related coordinate problem
- $feature->start($feature->start + VC_HACK_BP);
- $feature->end($feature->end + VC_HACK_BP);
- # store on overlap, unless we're not interested in this hit seq.
- foreach my $exon (@exons) {
- if ($exon->overlaps($feature)) {
- my $hid = $feature->hseqname;
- my $wanted = 0;
- if (@wanted_arr) {
- foreach (@wanted_arr) { # we only want some hits
- if ($_ eq $hid) {
- $wanted = 1;
- }
- }
- } else { # we want all hits
- $wanted = 1;
- }
- if ($wanted) {
- push @features, $feature;
- next FEATURE_LOOP;
- }
- }
- }
- }
- }
-
- my $nonredundant_f_arr_ref = $self->_remove_duplicate_features(
- \@features);
- return @$nonredundant_f_arr_ref;
-}
-
-# _remove_duplicate_features: takes ref to an array of features,
-# returns ref to an array of those features after duplicates removed
-
-sub _remove_duplicate_features {
- my ($self, $feature_arr_ref) = @_;
- $self->throw('interface fault') if (@_ != 2);
-
- my @sorted_features = sort { $a->hseqname cmp $b->hseqname
- || $a->start <=> $b->start
- || $a->end <=> $b->end
- || $a->hstart <=> $b->hstart
- || $a->hend <=> $b->hend
- || $a->strand <=> $b->strand
- } @$feature_arr_ref;
- for (my $i = 1; $i < @sorted_features; $i++) {
- my $f1 = $sorted_features[$i];
- my $f2 = $sorted_features[$i-1];
- if ( not $f1->hseqname cmp $f2->hseqname
- || $f1->start <=> $f2->start
- || $f1->end <=> $f2->end
- || $f1->hstart <=> $f2->hstart
- || $f1->hend <=> $f2->hend
- || $f1->strand <=> $f2->strand )
- {
- splice @sorted_features, $i, 1;
- $i--;
- }
- }
- return \@sorted_features;
-}
-
-=head2 _get_features_from_rawcontig
-
- Title : _get_features_from_rawcontig
- Usage : $ea->_get_features_from_rawcontig($rc_obj, $strand);
- $ea->_get_features_from_rawcontig($rc_obj, $strand,
- $logic_name);
- Function: Get features off specified strand of the raw contig
- supplied; if a list of logic names is given,
- features not resulting from an analysis with that
- logic name are cut; duplicate features are removed
- Returns : array of featurepairs
-
-=cut
-
-sub _get_features_from_rawcontig {
- my ($self, $rawcontig_obj, $strand, $db) = splice @_, 0, 4;
- $self->throw('interface fault') if (!$self or !$rawcontig_obj or !$strand
- or !$db);
- my @wanted_arr = @_;
-
- my $rca = $db->get_RawContigAdaptor;
-
- # keeping protein and DNA together for simplified porting
- # from branch-ensembl-6
- my $pfadp = $db->get_ProteinAlignFeatureAdaptor;
- my @gapped_features = @{$pfadp->fetch_by_Contig($rawcontig_obj)};
- my $dfadp = $db->get_DnaAlignFeatureAdaptor;
- push @gapped_features, @{$dfadp->fetch_by_Contig($rawcontig_obj)};
-
- my @all_features = ();
- GAPPED_FEATURE_LOOP:
- foreach my $gapped_feature (@gapped_features) {
- if (@wanted_arr) {
- my $wanted = 0;
- foreach (@wanted_arr) {
- if ($_ eq $gapped_feature->analysis->logic_name) {
- $wanted = 1;
- }
- }
- next GAPPED_FEATURE_LOOP if ! $wanted;
- }
- push @all_features, $gapped_feature->_parse_cigar;
- }
-
- my @features = ();
- foreach my $feature (@all_features) {
- next unless $feature->primary_tag =~ /similarity/i;
- if ($feature->strand == $strand) {
- if (! $feature->can('hseqname')) {
- $self->warn("feature $feature has no hseqname method");
- } else {
- push @features, $feature;
- }
- }
- }
- my $sorted_features_arr_ref = $self->_remove_duplicate_features(\@features);
- return @$sorted_features_arr_ref;
-}
-
-=head2 _get_Seqs_by_accs
-
- Title : _get_Seqs_by_accs
- Usage : $seqs = $self->_get_Seqs_by_accs(@hid_arr);
- Function: Does the sequence retrieval for an array of
- accesions in one pfetch call. This is closely
- based on
- Bio/EnsEMBL::Pipeline::SeqFetcher::Pfetch in
- ensembl-pipeline. Having it here allows the
- caller to simply specify a path to the pfetch
- executable (or use the default path). This
- allows us to avoid the different, incompatible
- wrappers for pfetch in the Web and pipeline
- code, and hence work with either pipeline or
- Web. However, if $self->seqfetcher does exist,
- that will be used in preference to the pfetch
- executable.
- Returns : Array of Bio::Seq
-
-=cut
-
-sub _get_Seqs_by_accs {
- my ($self, @acc) = @_;
-
- if (!@acc || scalar(@acc < 1)) {
- $self->throw("No accession input");
- }
-
- my @seq;
-
- if ($self->seqfetcher) {
- eval {
- @seq = $self->seqfetcher->get_Seqs_by_accs(@acc);
- };
- if($@){
- $self->warn("$@");
- }
- return @seq;
- }
-
- my $newseq;
- my $tracker = 0;
- my $pfetch = $self->pfetch;
-
- my $command = "$pfetch -q ";
- $command .= join " ", @acc;
-
- open(EVIDENCEALIGNMENT_IN_FH,"$command |")
- or $self->throw("Error opening pipe to pfetch for $pfetch");
- while(<EVIDENCEALIGNMENT_IN_FH>){
-
- chomp;
- eval{
- if(defined $_ && $_ ne "no match") {
- $newseq = new Bio::Seq('-seq' => $_,
- '-accession_number' => $acc[$tracker],
- '-display_id' => $acc[$tracker]);
- }
- };
-
- if($@){
- $self->warn("$@");
- }
-
- if (defined $newseq){
- push (@seq, $newseq);
- }
- else{
- $self->warn("Could not even pfetch sequence for [" . $acc[$tracker] . "]");
- }
- $tracker++;
- }
-
- close EVIDENCEALIGNMENT_IN_FH
- or $self->throw("Error running pfetch for $pfetch");
- return @seq;
-}
-
-=head2 _pad_pep_str
-
- Title : _pad_pep_str
- Usage : $padded_pep_seq = $ea->_pad_pep_str($pep_str);
- Function: make a protein sequence the same length as the
- corresponding DNA sequence by adding '--' after
- each amino acid
- Returns : string
-
-=cut
-
-sub _pad_pep_str {
- my ($self, $original) = @_;
- $self->throw('interface fault') if (@_ != 2);
-
- my @amino_acids = split //, $original;
- my $padded = '';
- foreach my $amino_acid (@amino_acids) {
- $padded .= $amino_acid . '--';
- }
- return $padded;
-}
-
-=head2 fetch_alignment
-
- Title : fetch_alignment
- Usage : my @seq_arr = $ea->fetch_alignment;
- my @seq_arr = $ea->fetch_alignment($hid);
- Function: gets transcript or raw contig and corresponding
- evidence or similarity features; for raw
- contigs, these are displayed for the forward
- strand followed by the reverse strand
- Args : for transcripts, an optional list of accession
- numbers of hit sequences of interest - if none
- are given, all relevant hit sequences are
- retrieved; for contigs, an optional list of
- logic names of hits of interest - if none
- are given, all hits are shown irrespective of
- logic name
- Returns : array of Bio::PrimarySeq
-
-=cut
-
-sub fetch_alignment {
- my ($self) = shift;
- $self->throw('interface fault') if (! $self);
- $self->throw('must have a transcript or contig ID and a DB adaptor object')
- unless (($self->transcriptid || $self->contigname) && $self->dbadaptor);
-
- my $ref_to_retval; # reference to array to return
- if ($self->transcriptid) {
- $ref_to_retval = $self->_get_aligned_evidence_for_transcript(
- $self->transcriptid, $self->dbadaptor, @_);
- } elsif ($self->contigname) {
- my $rca = $self->dbadaptor->get_RawContigAdaptor;
- my $contig = $rca->fetch_by_name($self->contigname);
- my $plus_strand_alignment = $self->_get_aligned_features_for_contig(
- $contig, $self->dbadaptor, 1, @_);
- my $minus_strand_alignment = $self->_get_aligned_features_for_contig(
- $contig, $self->dbadaptor, -1, @_);
- my $all_alignments = $plus_strand_alignment;
- foreach my $line (@$minus_strand_alignment) {
- push @$all_alignments, $line;
- }
- $ref_to_retval = $all_alignments;
- }
- return @$ref_to_retval;
-}
-
-# sort an evidence array
-sub _evidence_sort {
- my ($self, $evidence_arr_ref) = @_;
- $self->throw('interface fault') if (@_ != 2);
-
- # ENST... first. Then alphabetically by name, followed by indent
- my @sorted_evidence_arr = sort { ( $$a{hseqname} =~ /^ENST/ ? -1 : 0 )
- || ( $$b{hseqname} =~ /^ENST/ ? 1 : 0 )
- || ( $$a{hseqname} cmp $$b{hseqname} )
- || ( $$a{hindent} <=> $$b{hindent} )
- } @$evidence_arr_ref;
- return \@sorted_evidence_arr;
-}
-
-# get cDNA
-sub _get_transcript_nuc {
- my ($self, $exon_arr_ref) = @_;
- $self->throw('interface fault') if (@_ != 2);
-
- my $retval = '';
- my $seq_str;
- foreach my $exon (@$exon_arr_ref) {
- $retval .= $exon->seq->seq;
- }
- return $retval;
-}
-
-# get all hit sequences by aggregated use of pfetch
-sub _get_hits {
- my ($self, $features_arr_ref) = @_;
- $self->throw('interface fault') if (@_ != 2);
-
- my $clump_size = 1000; # number of sequences to fetch at once
- my %hits_hash = ();
- my @hseqnames = ();
- for (my $i = 0; $i < @$features_arr_ref; $i++) {
- my $hseqname = $$features_arr_ref[$i]->hseqname;
- if (! exists $hits_hash{$hseqname})
- {
- push @hseqnames, $hseqname;
- if ((@hseqnames % $clump_size) == 0) {
- my @tofetch = sort { $a cmp $b } @hseqnames;
- for (my $i = 1; $i < @tofetch; $i++) {
- if ($tofetch[$i] eq $tofetch[$i-1]) {
- splice @tofetch, $i, 1;
- $i--;
- }
- }
- for (my $i = 1; $i < @tofetch; $i++) {
- if ($tofetch[$i] =~ /\|/) { # pipe would make trouble
- splice @tofetch, $i, 1;
- $i--;
- }
- }
- my @seqs;
- if (@tofetch) {
- @seqs = $self->_get_Seqs_by_accs(@tofetch);
- }
- foreach my $seq_obj (@seqs) {
- $hits_hash{$seq_obj->accession_number} = $seq_obj;
- }
- foreach my $hseqname (@hseqnames) {
- if (not exists $hits_hash{$hseqname}) {
- $self->warn("couldn't fetch sequence for hit $hseqname");
- $hits_hash{$hseqname} = undef;
- }
- }
- @hseqnames = ();
- }
- }
- }
- if (@hseqnames) { # fetch the non-clump-sized remainder
- my @tofetch = sort { $a cmp $b } @hseqnames;
- for (my $i = 1; $i < @tofetch; $i++) {
- if ($tofetch[$i] eq $tofetch[$i-1]) {
- splice @tofetch, $i, 1;
- $i--;
- }
- }
- for (my $i = 1; $i < @tofetch; $i++) {
- if ($tofetch[$i] =~ /\|/) { # pipe would make trouble
- splice @tofetch, $i, 1;
- $i--;
- }
- }
- my @seqs;
- if (@tofetch) {
- @seqs = $self->_get_Seqs_by_accs(@tofetch);
- }
- foreach my $seq_obj (@seqs) {
- $hits_hash{$seq_obj->accession_number} = $seq_obj;
- }
- foreach my $hseqname (@hseqnames) {
- if (not exists $hits_hash{$hseqname}) {
- $self->warn("couldn't fetch sequence for hit $hseqname");
- $hits_hash{$hseqname} = undef;
- }
- }
- }
-
- return \%hits_hash;
-}
-
-# _evidence_lines_sort: takes reference to an array of evidence lines
-# and reference to a hash of scores,
-# returns reference to the former sorted by score (highest score first),
-# with ties sorted alphabetically
-
-sub _evidence_lines_sort {
- my ($self, $tmp_evidence_arr_ref, $scores_hash_ref) = @_;
- $self->throw('interface fault') if (@_ != 3);
-
- my @sorted_arr = sort {
- $$scores_hash_ref{$b->accession_number}
- <=> $$scores_hash_ref{$a->accession_number}
- || $a->accession_number cmp $b->accession_number
- } @$tmp_evidence_arr_ref;
-
- return \@sorted_arr;
-}
-
-# _get_per_hid_effective_scores: takes reference to an array of features,
-# returns a reference to a hash giving the 'effective score' for each
-# hseqname, i.e., the score-like value upon which we wish to sort.
-
-sub _get_per_hid_effective_scores {
- my ($self, $feature_arr_ref) = @_;
- $self->throw('interface fault') if (@_ != 2);
-
- my %per_hid_effective_scores = ();
- foreach my $feature (@$feature_arr_ref) {
- my $hseqname = $feature->hseqname;
- my $feature_len = $feature->end - $feature->start + 1;
- if (not exists $per_hid_effective_scores{$hseqname})
- {
- $per_hid_effective_scores{$hseqname} = 0;
- }
- $per_hid_effective_scores{$hseqname} += $feature_len;
- }
- return \%per_hid_effective_scores;
-}
-
-# _get_aligned_features_for_contig: takes a contig object, a DB adaptor
-# and strand
-# returns ref to an array of Bio::PrimarySeq
-
-sub _get_aligned_features_for_contig {
- my ($self, $contig_obj, $db, $strand) = splice @_, 0, 4;
- $self->throw('interface fault') if (!$self or !$contig_obj or !$db
- or ! $strand);
-
- my @evidence_arr; # a reference to this is returned
- my $evidence_obj;
-
- # get contig
- my $rca = $db->get_RawContigAdaptor;
-
- my @features = $self->_get_features_from_rawcontig($contig_obj,
- $strand, $db, @_);
- my $per_hid_effective_scores_hash_ref
- = $self->_get_per_hid_effective_scores(\@features);
- my $hits_hash_ref = $self->_get_hits(\@features);
-
- my $nucseq_obj = $contig_obj;
-
- if ($strand < 0) {
- $nucseq_obj = $nucseq_obj->revcom;
- }
- my $nucseq_str = $nucseq_obj->seq;
- my $dna_len_bp = length $nucseq_str;
- my @translations = ();
- for (my $i = 0; $i < 3; $i++) {
- push @translations, $nucseq_obj->translate(undef, undef, $i);
- }
-
- # protein evidence
-
- # translations themselves form our first three rows of 'evidence'
- for (my $i = 0; $i < 3; $i++) {
- my $evidence_line = $translations[$i]->seq;
- $evidence_line = $self->_pad_pep_str($evidence_line);
- if ($i == 1) {
- $evidence_line = '-' . $evidence_line;
- } elsif ($i == 2) {
- $evidence_line = '--' . $evidence_line;
- }
- while (length $evidence_line < $dna_len_bp) {
- $evidence_line .= '-';
- }
- while (length $evidence_line > $dna_len_bp) {
- chop $evidence_line;
- }
- $evidence_obj = Bio::PrimarySeq->new(
- -seq => $evidence_line,
- -id => 0,
- -accession_number => $contig_obj->name,
- -moltype => 'protein'
- );
- push @evidence_arr, $evidence_obj;
- }
-
- my @pep_evidence_arr = ();
- my $last_feat = undef;
-
- PEP_FEATURE_LOOP:
- foreach my $feature (@features) {
- if (($feature->start < 1) || ($feature->end > $dna_len_bp)) {
- $self->warn("genomic coordinates out of range: start " .
- $feature->start . ", end " . $feature->end);
- next PEP_FEATURE_LOOP;
- }
- my $hit_seq_obj = $$hits_hash_ref{$feature->hseqname};
- if (! $hit_seq_obj) {
- next PEP_FEATURE_LOOP; # already warned in _get_hits()
- }
- next PEP_FEATURE_LOOP # not an error, DNA and protein are mixed
- unless ($hit_seq_obj->moltype eq 'protein');
- my $hlen = $feature->hend - $feature->hstart + 1;
- my $flen = $feature->end - $feature->start + 1;
- if ($flen != 3 * $hlen) {
- $self->warn("genomic length $flen but protein hit length $hlen for hit "
- . $feature->hseqname);
- next PEP_FEATURE_LOOP;
- }
- if (($feature->hstart - 1 < 0) || ($feature->hstart - 1 + $hlen
- > length $hit_seq_obj->seq))
- {
- $self->warn("hit coordinates out of range: hit " . $feature->hseqname .
- ", hit start " . $feature->hstart . ", hit length $hlen");
- next PEP_FEATURE_LOOP;
- }
- my $hseq = substr $hit_seq_obj->seq, $feature->hstart - 1, $hlen;
- $hseq = $self->_pad_pep_str($hseq);
- my $hindent_bp;
- if ($strand > 0) {
- $hindent_bp = $feature->start - 1;
- } else {
- $hindent_bp = $dna_len_bp - $feature->end;
- }
- if ($hindent_bp < 0) {
- $hindent_bp = 0; # disaster recovery
- }
- my %hit_details = ( 'moltype' => $hit_seq_obj->moltype,
- 'hseqname' => $feature->hseqname,
- 'hseq' => $hseq,
- 'hindent' => $hindent_bp,
- );
- push @pep_evidence_arr, \%hit_details;
- $last_feat = $feature;
- }
-
- my @sorted_pep_evidence_arr = @{$self->_evidence_sort(\@pep_evidence_arr)};
- my @tmp_pep_evidence_arr = ();
-
- my $evidence_line = '';
- my $prev_hseqname = '-' x 1000; # fake initial ID
- for (my $i = 0; $i < @sorted_pep_evidence_arr; $i++) {
- my $hit = $sorted_pep_evidence_arr[$i];
- if ($$hit{hseqname} ne $prev_hseqname) { # make new evidence line
- $evidence_line = '-' x $dna_len_bp;
- }
-
- # splice in the evidence fragment
- my $hseqlen = length $$hit{hseq};
- next
- if (($$hit{hindent} < 0) || ($$hit{hindent} + $hseqlen > $dna_len_bp));
- substr $evidence_line, $$hit{hindent}, $hseqlen, $$hit{hseq};
-
- # store if end of evidence line
- if (($i == $#sorted_pep_evidence_arr)
- || ($sorted_pep_evidence_arr[$i+1]{hseqname} ne $$hit{hseqname}))
- {
- $evidence_obj = Bio::PrimarySeq->new(
- -seq => $evidence_line,
- -id => 0,
- -accession_number => $$hit{hseqname},
- -moltype => $$hit{moltype}
- );
- push @tmp_pep_evidence_arr, $evidence_obj;
- }
- $prev_hseqname = $$hit{hseqname};
- }
-
- my $sorted_pep_evidence_lines_ref =
- $self->_evidence_lines_sort(\@tmp_pep_evidence_arr,
- $per_hid_effective_scores_hash_ref);
- push @evidence_arr, @$sorted_pep_evidence_lines_ref;
-
- # nucleic acid evidence
-
- # DNA itself forms first row of nucleic acid evidence
- $evidence_obj = Bio::PrimarySeq->new(
- -seq => $nucseq_str,
- -id => 0,
- -accession_number => $contig_obj->name,
- -moltype => 'dna'
- );
- push @evidence_arr, $evidence_obj;
-
- my @nuc_evidence_arr = ();
- $last_feat = undef;
- NUC_FEATURE_LOOP:
- foreach my $feature(@features) {
- if (($feature->start < 1) || ($feature->end > $dna_len_bp)) {
- $self->warn("genomic coordinates out of range: start " .
- $feature->start . ", end " . $feature->end);
- next NUC_FEATURE_LOOP;
- }
- my $hit_seq_obj = $$hits_hash_ref{$feature->hseqname};
- if (! $hit_seq_obj) {
- next NUC_FEATURE_LOOP; # already warned in _get_hits()
- }
- next NUC_FEATURE_LOOP # not an error, DNA and protein are mixed
- unless ($hit_seq_obj->moltype ne 'protein');
- my $hlen = $feature->hend - $feature->hstart + 1;
- my $flen = $feature->end - $feature->start + 1;
- if ($hlen != $flen) {
- $self->warn("genomic length $flen but DNA hit length $hlen for hit "
- . $feature->hseqname);
- next NUC_FEATURE_LOOP;
- }
- if (($feature->hstart - 1 < 0) || ($feature->hstart - 1 + $hlen
- > length $hit_seq_obj->seq))
- {
- $self->warn("hit coordinates out of range: hit " . $feature->hseqname .
- ", hit start " . $feature->hstart . ", hit length $hlen");
- next NUC_FEATURE_LOOP;
- }
- my $hseq = substr $hit_seq_obj->seq, $feature->hstart - 1, $hlen;
- my $strand_wrt_dna = $strand * $feature->strand;
- if ($strand_wrt_dna < 0) { # reverse-compliment the hit
- my $hseq_obj_tmp = Bio::PrimarySeq->new(
- -seq => $hseq,
- -id => 'fake_id',
- -accession_number => '0',
- -moltype => $hit_seq_obj->moltype
- );
- $hseq = $hseq_obj_tmp->revcom->seq;
- }
- my $hindent_bp;
- if ($strand > 0) {
- $hindent_bp = $feature->start - 1;
- } else{
- $hindent_bp = $dna_len_bp - $feature->end;
- }
- if ($hindent_bp < 0) {
- $hindent_bp = 0; # disaster recovery
- }
- my %hit_details = ( 'moltype' => $hit_seq_obj->moltype,
- 'hseqname' => $feature->hseqname,
- 'hseq' => $hseq,
- 'hindent' => $hindent_bp,
- );
- push @nuc_evidence_arr, \%hit_details;
- $last_feat = $feature;
- }
-
- my @sorted_nuc_evidence_arr = @{$self->_evidence_sort(\@nuc_evidence_arr)};
- my @tmp_nuc_evidence_arr = ();
-
- $evidence_line = '';
- my $hit = $sorted_nuc_evidence_arr[0];
- $prev_hseqname = '-' x 1000; # fake initial ID
- for (my $i = 0; $i < @sorted_nuc_evidence_arr; $i++) {
- my $hit = $sorted_nuc_evidence_arr[$i];
- my $hseq_str = $$hit{hseq};
- if ($$hit{hseqname} ne $prev_hseqname) { # make new evidence line
- $evidence_line = '-' x $dna_len_bp;
- }
-
- # splice in the evidence fragment
- my $hseqlen = length $$hit{hseq};
- next unless ($$hit{hindent} + $hseqlen <= $dna_len_bp);
- substr $evidence_line, $$hit{hindent}, $hseqlen, $hseq_str;
-
- # store if end of evidence line
- if (($i == $#sorted_nuc_evidence_arr)
- || ($sorted_nuc_evidence_arr[$i+1]{hseqname} ne $$hit{hseqname}))
- {
- $evidence_obj = Bio::PrimarySeq->new(
- -seq => $evidence_line,
- -id => 0,
- -accession_number => $$hit{hseqname},
- -moltype => $$hit{moltype}
- );
- push @tmp_nuc_evidence_arr, $evidence_obj;
- }
- $prev_hseqname = $$hit{hseqname};
- }
-
- my $sorted_nuc_evidence_lines_ref =
- $self->_evidence_lines_sort(\@tmp_nuc_evidence_arr,
- $per_hid_effective_scores_hash_ref);
- push @evidence_arr, @$sorted_nuc_evidence_lines_ref;
-
- # remove blank evidence lines
-
- my @filtered_evidence_arr = ();
- foreach my $evidence_line (@evidence_arr) {
- push @filtered_evidence_arr, $evidence_line
- if ($$evidence_line{seq} =~ /[^-]/);
- }
-
- # use uppercase
- foreach my $evidence_line (@filtered_evidence_arr) {
- my $seq_str = uc $evidence_line->seq;
- $evidence_line->seq($seq_str);
- }
-
- return \@filtered_evidence_arr;
-}
-
-# _get_aligned_evidence_for_transcript: takes a transcript ID and
-# a DB adaptor, and an optional list of hit sequence accession numbers
-# returns ref to an array of Bio::PrimarySeq
-
-sub _get_aligned_evidence_for_transcript {
- my ($self, $transcript_id, $db) = splice @_, 0, 3;
- $self->throw('interface fault')
- if (!$self or !$transcript_id or !$db);
-
- my $sa = $db->get_SliceAdaptor;
- my @evidence_arr; # a reference to this is returned
- my $evidence_obj;
-
- my $ta = $db->get_TranscriptAdaptor;
- my $ga = $db->get_GeneAdaptor;
- my $ea = $db->get_ExonAdaptor;
- my $transcript_name_to_display = $transcript_id;
-
- # get all exons off a slice
- my $transcript_obj_nonslice;
- my $transcript_dbID;
- if ($transcript_id =~ /^ENS/i) { # stable ID
- $transcript_obj_nonslice = $ta->fetch_by_stable_id($transcript_id);
- $transcript_dbID = $transcript_obj_nonslice->dbID;
- } else { # internal (db) ID
- $transcript_dbID = $transcript_id;
- $transcript_obj_nonslice = $ta->fetch_by_dbID($transcript_dbID);
- }
- my $slice = $sa->fetch_by_transcript_id($transcript_dbID,
- 1000);
-
- my $transcript_obj; # VC version
- my $genes = $slice->get_all_Genes;
- GENE_LOOP:
- foreach my $gene (@$genes) {
- my $transcripts = $gene->get_all_Transcripts;
- foreach my $transcript (@$transcripts) {
- if ($transcript->dbID eq $transcript_dbID) {
- $transcript_obj = $transcript;
- last GENE_LOOP;
- }
- }
- }
- my @all_exons = @{$transcript_obj->get_all_Exons};
-
- my @features;
- if ($self->use_supporting_evidence) {
- @features = $self->_get_evidence_from_transcript(
- $transcript_obj, @_);
- } else {
- @features = $self->_get_similarity_features_from_transcript(
- $transcript_obj, $slice, @_);
- }
- my $per_hid_effective_scores_hash_ref =
- $self->_get_per_hid_effective_scores(\@features);
- my $hits_hash_ref = $self->_get_hits(\@features);
- my $nucseq_str = $self->_get_transcript_nuc(\@all_exons);
- my $cdna_len_bp = length $nucseq_str;
-
- # protein evidence
-
- my @utr_free_exons = $transcript_obj->translateable_exons;
-
- # record whether each exon is totally UTR, simply a frameshift (hence
- # not to be translated), or at least partly translating
- my @exon_status = ();
- for (my $i = 0; $i < @all_exons; $i++) {
- $exon_status[$i] = 'utr';
- }
- for (my $i = 0; $i < @all_exons; $i++) {
- if (($all_exons[$i]->end - $all_exons[$i]->start + 1) < 3) {
- $exon_status[$i] = 'frameshift';
- } else {
- foreach my $utr_free_exon (@utr_free_exons) {
- if (($utr_free_exon->start >= $all_exons[$i]->start)
- && ($utr_free_exon->end <= $all_exons[$i]->end)) {
- $exon_status[$i] = 'translating';
- }
- }
- }
- }
-
- # get length of the UTR witin the first exon (excludes length of any
- # exons that fall entirely within the 5' UTR)
- # and also get the total 5' UTR length
- my $total_5prime_utr_len = 0;
- my $fiveprime_utr_in_first_exon = 0;
- UTR_OUTER:
- for (my $i = 0; $i < @all_exons; $i++) {
- my $exon_i = $all_exons[$i];
- if ($exon_status[$i] ne 'translating') {
- $total_5prime_utr_len += $exon_i->end - $exon_i->start + 1;
- } else { # 1st translating exon, could contain part of 5' UTR
- foreach my $utr_free_exon (@utr_free_exons) {
- if (($utr_free_exon->start >= $exon_i->start)
- && ($utr_free_exon->end <= $exon_i->end))
- {
- if ($exon_i->strand > 0) {
- $fiveprime_utr_in_first_exon = $utr_free_exon->start
- - $exon_i->start;
- } else {
- $fiveprime_utr_in_first_exon = $exon_i->end - $utr_free_exon->end;
- }
- last UTR_OUTER;
- }
- }
- }
- }
- if ($fiveprime_utr_in_first_exon < 0) {
- $fiveprime_utr_in_first_exon = 0; # disaster recovery
- }
- $total_5prime_utr_len += $fiveprime_utr_in_first_exon;
-
- # get 3' UTR length
- my $total_3prime_utr_len = $cdna_len_bp
- - length($transcript_obj->translate->seq) * 3
- - $total_5prime_utr_len;
- if ($total_3prime_utr_len < 0) {
- $total_3prime_utr_len = 0; # disaster recovery
- }
-
- # get translation, avoiding transcript's translate method due to rare problem
- my $seq_to_translate = substr $nucseq_str, $total_5prime_utr_len;
- my $cdna_obj = Bio::PrimarySeq->new(
- -seq => $seq_to_translate,
- -id => 0,
- -accession_number => $transcript_name_to_display,
- -moltype => 'dna'
- );
- my $translation_including_3prime_utr = $cdna_obj->translate->seq;
-
- # translation itself forms our first row of 'evidence'
- my $evidence_line = $translation_including_3prime_utr;
- $evidence_line = $self->_pad_pep_str($evidence_line);
- # adjust evidence line for UTRs
- $evidence_line = ('-' x $total_5prime_utr_len) . $evidence_line;
- $evidence_line = substr $evidence_line, 0, $cdna_len_bp
- - $total_3prime_utr_len;
- $evidence_line .= '-' x $total_3prime_utr_len;
- # store
- $evidence_obj = Bio::PrimarySeq->new(
- -seq => $evidence_line,
- -id => 0,
- -accession_number => $transcript_name_to_display,
- -moltype => 'protein'
- );
- push @evidence_arr, $evidence_obj;
-
- my $total_exon_len = 0;
- my @pep_evidence_arr = ();
- my $last_feat = undef;
- foreach my $exon (@all_exons) {
- PEP_FEATURE_LOOP:
- foreach my $feature(@features) {
- next PEP_FEATURE_LOOP # unless feature falls within this exon
- unless $exon->overlaps($feature);
- my $hit_seq_obj = $$hits_hash_ref{$feature->hseqname};
- if (! $hit_seq_obj) {
- next PEP_FEATURE_LOOP; # already warned in _get_hits()
- }
- next PEP_FEATURE_LOOP # not an error, DNA and protein are mixed
- unless ($hit_seq_obj->moltype eq 'protein');
- my $hlen = $feature->hend - $feature->hstart + 1;
- my $flen = $feature->end - $feature->start + 1;
- if ($flen != 3 * $hlen) {
- $self->warn("genomic length $flen but protein hit length $hlen for hit "
- . $feature->hseqname);
- next PEP_FEATURE_LOOP;
- }
- if (($feature->hstart - 1 < 0) || ($feature->hstart - 1 + $hlen
- > length $hit_seq_obj->seq))
- {
- $self->warn("hit coordinates out of range: hit " . $feature->hseqname .
- ", hit start " . $feature->hstart . ", hit length $hlen");
- next PEP_FEATURE_LOOP;
- }
- my $hseq = substr $hit_seq_obj->seq, $feature->hstart - 1, $hlen;
- $hseq = $self->_pad_pep_str($hseq);
- if ($feature->start < $exon->start) {
- my $old_flen = $flen;
- $flen -= $exon->start - $feature->start;
- $feature->start($exon->start);
- if ($exon->strand > 0) { # trim start of hit
- $hseq = substr $hseq, $flen - $old_flen, $flen;
- } else { # trim end of hit
- $hseq = substr $hseq, 0, $flen;
- }
- }
- if ($feature->end > $exon->end) {
- my $old_flen = $flen;
- $flen -= $feature->end - $exon->end;
- $feature->end($exon->end);
- if ($exon->strand > 0) { # trim end of hit
- $hseq = substr $hseq, 0, $flen;
- } else { # trim start of hit
- $hseq = substr $hseq, $old_flen - $flen, $flen;
- }
- }
- my $hindent_bp;
- if ($exon->strand > 0) {
- $hindent_bp = $total_exon_len + $feature->start - $exon->start;
- } else {
- $hindent_bp = $total_exon_len + $exon->end - $feature->end;
- }
- if ($hindent_bp < 0) {
- $hindent_bp = 0; # disaster recovery
- }
- my %hit_details = ( 'moltype' => $hit_seq_obj->moltype,
- 'hseqname' => $feature->hseqname,
- 'hseq' => $hseq,
- 'hindent' => $hindent_bp,
- 'exon' => $exon
- );
- push @pep_evidence_arr, \%hit_details;
- $last_feat = $feature;
- }
- $total_exon_len += $exon->end - $exon->start + 1;
- }
-
- my @sorted_pep_evidence_arr = @{$self->_evidence_sort(\@pep_evidence_arr)};
- my @tmp_pep_evidence_arr = ();
-
- $evidence_line = '';
- my $prev_hseqname = '-' x 1000; # fake initial ID
- for (my $i = 0; $i < @sorted_pep_evidence_arr; $i++) {
- my $hit = $sorted_pep_evidence_arr[$i];
- if ($$hit{hseqname} ne $prev_hseqname) { # make new evidence line
- $evidence_line = '-' x $cdna_len_bp;
- }
-
- # splice in the evidence fragment
- my $hseqlen = length $$hit{hseq};
- substr $evidence_line, $$hit{hindent}, $hseqlen, $$hit{hseq};
-
- # store if end of evidence line
- if (($i == $#sorted_pep_evidence_arr)
- || ($sorted_pep_evidence_arr[$i+1]{hseqname} ne $$hit{hseqname}))
- {
- # purge the UTRs of protein 'evidence'
- substr $evidence_line, 0, $total_5prime_utr_len,
- ('-' x $total_5prime_utr_len);
- substr $evidence_line, $cdna_len_bp - $total_3prime_utr_len,
- $total_3prime_utr_len, ('-' x $total_3prime_utr_len);
- $evidence_obj = Bio::PrimarySeq->new(
- -seq => $evidence_line,
- -id => 0,
- -accession_number => $$hit{hseqname},
- -moltype => $$hit{moltype}
- );
- push @tmp_pep_evidence_arr, $evidence_obj;
- }
- $prev_hseqname = $$hit{hseqname};
- }
-
- my $sorted_pep_evidence_lines_ref =
- $self->_evidence_lines_sort(\@tmp_pep_evidence_arr,
- $per_hid_effective_scores_hash_ref);
- push @evidence_arr, @$sorted_pep_evidence_lines_ref;
-
- # nucleic acid evidence
-
- # cDNA itself forms first row of nucleic acid evidence
- $evidence_obj = Bio::PrimarySeq->new(
- -seq => $nucseq_str,
- -id => 0,
- -accession_number => $transcript_name_to_display,
- -moltype => 'dna'
- );
- push @evidence_arr, $evidence_obj;
-
- $total_exon_len = 0;
- my @nuc_evidence_arr = ();
- $last_feat = undef;
- foreach my $exon (@all_exons) {
- NUC_FEATURE_LOOP:
- foreach my $feature(@features) {
- next NUC_FEATURE_LOOP # unless feature falls within this exon
- unless $exon->overlaps($feature);
- my $hit_seq_obj = $$hits_hash_ref{$feature->hseqname};
- if (! $hit_seq_obj) {
- next NUC_FEATURE_LOOP; # already warned in _get_hits()
- }
- next NUC_FEATURE_LOOP # not an error, DNA and protein are mixed
- unless ($hit_seq_obj->moltype ne 'protein');
- my $hlen = $feature->hend - $feature->hstart + 1;
- my $flen = $feature->end - $feature->start + 1;
- if ($hlen != $flen) {
- $self->warn("genomic length $flen but DNA hit length $hlen for hit "
- . $feature->hseqname);
- next NUC_FEATURE_LOOP;
- }
- if (($feature->hstart - 1 < 0) || ($feature->hstart - 1 + $hlen
- > length $hit_seq_obj->seq))
- {
- $self->warn("hit coordinates out of range: hit " . $feature->hseqname .
- ", hit start " . $feature->hstart . ", hit length $hlen");
- next NUC_FEATURE_LOOP;
- }
- if ($feature->start < $exon->start) {
- my $old_flen = $flen;
- $flen -= $exon->start - $feature->start;
- $hlen -= $exon->start - $feature->start;
- $feature->start($exon->start);
- if ($exon->strand > 0) { # trim start of hit
- $feature->hstart($feature->hstart + $old_flen - $flen);
- } else { # trim end of hit
- $feature->hend($feature->hend - ($old_flen - $flen));
- }
- }
- if ($feature->end > $exon->end) {
- my $old_flen = $flen;
- $flen -= $feature->end - $exon->end;
- $hlen -= $feature->end - $exon->end;
- $feature->end($exon->end);
- if ($exon->strand > 0) { # trim end of hit
- $feature->hend($feature->hend - ($old_flen - $flen));
- } else { # trim start of hit
- $feature->hstart($feature->hstart + $old_flen - $flen);
- }
- }
- my $hseq = substr $hit_seq_obj->seq, $feature->hstart - 1, $hlen;
- my $strand_wrt_exon = $exon->strand * $feature->strand;
- if ($strand_wrt_exon < 0) { # reverse-compliment the hit
- my $hseq_obj_tmp = Bio::PrimarySeq->new(
- -seq => $hseq,
- -id => 'fake_id',
- -accession_number => '0',
- -moltype => $hit_seq_obj->moltype
- );
- $hseq = $hseq_obj_tmp->revcom->seq;
- }
- my $hindent_bp;
- if ($exon->strand > 0) {
- $hindent_bp = $total_exon_len + $feature->start - $exon->start;
- } else{
- $hindent_bp = $total_exon_len + $exon->end - $feature->end;
- }
- if ($hindent_bp < 0) {
- $hindent_bp = 0; # disaster recovery
- }
- my %hit_details = ( 'moltype' => $hit_seq_obj->moltype,
- 'hseqname' => $feature->hseqname,
- 'hseq' => $hseq,
- 'hindent' => $hindent_bp,
- 'exon' => $exon
- );
- push @nuc_evidence_arr, \%hit_details;
- $last_feat = $feature;
- }
- $total_exon_len += $exon->end - $exon->start + 1;
- }
-
- my @sorted_nuc_evidence_arr = @{$self->_evidence_sort(\@nuc_evidence_arr)};
- my @tmp_nuc_evidence_arr = ();
-
- $evidence_line = '';
- my $hit = $sorted_nuc_evidence_arr[0];
- my $prev_exon = $$hit{exon};
- $prev_hseqname = '-' x 1000; # fake initial ID
- for (my $i = 0; $i < @sorted_nuc_evidence_arr; $i++) {
- my $hit = $sorted_nuc_evidence_arr[$i];
- my $hseq_str = $$hit{hseq};
- if ($$hit{hseqname} ne $prev_hseqname) { # make new evidence line
- $evidence_line = '-' x $cdna_len_bp;
- }
-
- # splice in the evidence fragment
- my $hseqlen = length $$hit{hseq};
- next unless ($$hit{hindent} + $hseqlen <= $cdna_len_bp);
- substr $evidence_line, $$hit{hindent}, $hseqlen, $hseq_str;
-
- # store if end of evidence line
- if (($i == $#sorted_nuc_evidence_arr)
- || ($sorted_nuc_evidence_arr[$i+1]{hseqname} ne $$hit{hseqname}))
- {
- $evidence_obj = Bio::PrimarySeq->new(
- -seq => $evidence_line,
- -id => 0,
- -accession_number => $$hit{hseqname},
- -moltype => $$hit{moltype}
- );
- push @tmp_nuc_evidence_arr, $evidence_obj;
- }
- $prev_hseqname = $$hit{hseqname};
- $prev_exon = $$hit{exon};
- }
-
- my $sorted_nuc_evidence_lines_ref =
- $self->_evidence_lines_sort(\@tmp_nuc_evidence_arr,
- $per_hid_effective_scores_hash_ref);
- push @evidence_arr, @$sorted_nuc_evidence_lines_ref;
-
- # remove blank evidence lines
-
- my @filtered_evidence_arr = ();
- foreach my $evidence_line (@evidence_arr) {
- push @filtered_evidence_arr, $evidence_line
- if ($$evidence_line{seq} =~ /[^-]/);
- }
-
- # for transcripts: remove cDNA if one protein hid specified,
- # remove translation if one nucleotide hid specified, but (in
- # desperation) leave translation in place if there are no hits
- # of any kind
-
- if ($self->transcriptid) {
- my $prot_lines = 0;
- my $nuc_lines = 0;
- foreach my $evidence_line (@filtered_evidence_arr) {
- if ($evidence_line->moltype eq 'protein') {
- $prot_lines++;
- } else {
- $nuc_lines++;
- }
- }
- if ($nuc_lines eq 1) { # for nucleotides, we have cDNA only
- splice @filtered_evidence_arr, $#filtered_evidence_arr, 1;
- }
- if (@filtered_evidence_arr > 1 and $prot_lines == 1) {
- splice @filtered_evidence_arr, 0, 1;
- }
- }
-
- # change case at exon boundaries
-
- my @exon_lengths = ();
- foreach my $exon (@all_exons) {
- push @exon_lengths, ($exon->end - $exon->start + 1);
- }
- foreach my $evidence_line (@filtered_evidence_arr) {
- my $seq_str = $evidence_line->seq;
- $total_exon_len = 0;
- for (my $i = 0; $i < @exon_lengths; $i++) {
- my $replacement = substr $seq_str, $total_exon_len,
- $exon_lengths[$i];
- if ($i % 2) {
- $replacement = lc $replacement;
- } else {
- $replacement = uc $replacement;
- }
- substr $seq_str, $total_exon_len, $exon_lengths[$i], $replacement;
- $total_exon_len += $exon_lengths[$i];
- }
- $evidence_line->seq($seq_str);
- }
-
- return \@filtered_evidence_arr;
-}
-
-1;