From 96f7b6c6d9259772da37286f31d90e8fa6a5cd7d Mon Sep 17 00:00:00 2001
From: Kieron Taylor <ktaylor@ebi.ac.uk>
Date: Fri, 28 Oct 2011 15:28:32 +0000
Subject: [PATCH] Introduced Utils::IO namespace along with serializer code for
GFF3 and human-readable formats. Pre-release, not intended for public use.
Also added Utils::BiotypeMapper that performs ontology mappings.
---
modules/Bio/EnsEMBL/Utils/BiotypeMapper.pm | 221 ++++++++++++++++++
modules/Bio/EnsEMBL/Utils/IO/GFFSerializer.pm | 214 +++++++++++++++++
.../Bio/EnsEMBL/Utils/IO/ReportSerializer.pm | 209 +++++++++++++++++
modules/Bio/EnsEMBL/Utils/IO/Serializer.pm | 160 +++++++++++++
4 files changed, 804 insertions(+)
create mode 100644 modules/Bio/EnsEMBL/Utils/BiotypeMapper.pm
create mode 100644 modules/Bio/EnsEMBL/Utils/IO/GFFSerializer.pm
create mode 100644 modules/Bio/EnsEMBL/Utils/IO/ReportSerializer.pm
create mode 100644 modules/Bio/EnsEMBL/Utils/IO/Serializer.pm
diff --git a/modules/Bio/EnsEMBL/Utils/BiotypeMapper.pm b/modules/Bio/EnsEMBL/Utils/BiotypeMapper.pm
new file mode 100644
index 0000000000..d5148f0053
--- /dev/null
+++ b/modules/Bio/EnsEMBL/Utils/BiotypeMapper.pm
@@ -0,0 +1,221 @@
+=pod
+
+=head1 LICENSE
+
+ Copyright (c) 1999-2011 The European Bioinformatics Institute and
+ Genome Research Limited. All rights reserved.
+
+ This software is distributed under a modified Apache license.
+ For license details, please see
+
+ http://www.ensembl.org/info/about/code_licence.html
+
+=head1 NAME
+
+BiotypeMapper - Translates EnsEMBL biotypes into Sequence Ontology terms and back
+
+=head1 AUTHOR
+
+Kieron Taylor, 2011 - ktaylor@ebi.ac.uk
+
+=head1 SYNOPSIS
+
+use Bio::EnsEMBL::Utils::BiotypeMapper
+
+my $ontology_adaptor = $registry->get_adaptor( 'Multi', 'Ontology', 'OntologyTerm' );
+my $biotype_mapper = new BiotypeMapper($ontology_adaptor);
+
+print $biotype_mapper->translate_feature_to_SO_term($feature);
+
+=head1 DESCRIPTION
+
+BiotypeMapper provides a series of nearest matches between EnsEMBL biotypes and
+the Sequence Ontology (http://www.sequenceontology.org)
+
+Mappings are imperfect due to the inexact correspondance of biotypes to
+several SO terms. The a best guess has been chosen in each case.
+
+Reverse mappings from SO to biotype are vague, due to many-to-one relationships.
+In this case a list of possible terms is given.
+
+=cut
+
+package Bio::EnsEMBL::Utils::BiotypeMapper;
+
+use strict;
+use warnings;
+use Carp;
+
+my %gene_so_mapping = (
+ 'protein_coding' => 'SO:0001217', # protein_coding_gene
+ 'pseudogene' => 'SO:0000336', # pseudogene
+ 'processed_transcript' => 'SO:0001503', # processed_transcript
+ 'lincRNA' => 'SO:0001641', # lincRNA_gene
+ 'polymorphic_pseudogene'=> 'SO:0000336', # pseudogene
+ 'Mt_tRNA' => 'SO:0000088', # mt_gene
+ 'IG_D_gene' => 'SO:0000510', # D_gene
+ 'snoRNA' => 'SO:0001267', #snoRNA_gene
+ 'misc_RNA' => 'SO:0000356', #RNA
+ 'miRNA' => 'SO:0001265', #miRNA_gene
+ 'rRNA' => 'SO:0001637', #rRNA_gene
+ 'snRNA' => 'SO:0001268', #snRNA_gene
+ 'snRNA_pseudogene' => 'SO:0000336', # pseudogene
+ 'tRNA_pseudogene' => 'SO:0000778', # pseudogenic_tRNA
+ 'rRNA_pseudogene' => 'SO:0000777', # pseudogenic_rRNA
+ 'TR_J_gene' => 'SO:0000470', # J_gene
+ 'TR_V_gene' => 'SO:0000466', # V_gene
+ 'TR_C_gene' => 'SO:0000478', # C_gene
+ 'ncRNA' => 'SO:0001263', # ncRNA_gene
+ 'tRNA' => 'SO:0001272', # tRNA_gene
+ 'retrotransposed' => 'SO:0000569', # retrotransposed
+## heavily abbreviated
+ );
+
+my %transcript_so_mapping = (
+ 'processed_transcript' => 'SO:0001503', # processed_transcript
+ 'nonsense_mediated_decay' => 'SO:0001621', # NMD_transcript_variant
+ 'retained_intron' => 'SO:0000681', # aberrant_processed_transcript
+ 'transcribed_unprocessed_pseudogene'=> 'SO:0000516', # pseudogenic_transcript
+ 'processed_pseudogene' => 'SO:0000043', # processed_pseudogene
+ 'unprocessed_pseudogene' => 'SO:0000336', # pseudogene
+ 'unitary_pseudogene' => 'SO:0000336',
+ 'pseudogene' => 'SO:0000336', # pseudogene
+ 'transcribed_processed_pseudogene' => 'SO:0000043',
+ 'retrotransposed' => 'SO:0000569', #retrotransposed
+ 'ncrna_host' => 'SO:0000483',
+ 'polymorphic_pseudogene' => 'SO:0000336',
+ 'lincRNA' => 'SO:0001463',
+ 'ncrna_host' => 'SO:0000483',
+ '3prime_overlapping_ncrna' => 'SO:0000483',
+ 'TR_V_gene' => 'SO:0000466',
+ 'TR_V_pseudogene' => 'SO:0000336',
+
+ 'TR_J_gene' => 'SO:0000470',
+ 'IG_C_gene' => 'SO:0000478',
+ 'IG_C_pseudogene' => 'SO:0000336',
+ 'TR_C_gene' => 'SO:0000478',
+ 'IG_J_pseudogene' => 'SO:0000336',
+ 'miRNA' => 'SO:0000276', #miRNA
+ 'miRNA_pseudogene' => 'SO:0000336',
+ 'disrupted_domain' => 'SO:0000681', # aberrant_processed_transcript
+ 'rRNA' => 'SO:0000252', #rRNA
+ 'rRNA_pseudogene' => 'SO:0000777',
+ 'scRNA_pseudogene' => 'SO:0000336',
+ 'snoRNA' => 'SO:0000275', # snoRNA
+ 'snoRNA_pseudogene' => 'SO:0000336',
+ 'snRNA' => 'SO:0000274', # snRNA
+ 'snRNA_pseudogene' => 'SO:0000336',
+
+ );
+
+my %feature_so_mapping = (
+ 'Bio::EnsEMBL::Gene' => 'SO:0000704', # gene
+ 'Bio::EnsEMBL::Transcript' => 'SO:0000673', # transcript
+ 'Bio::EnsEMBL::Slice' => 'SO:0000001', # region
+ 'Bio::EnsEMBL::Variation::VariationFeature' => 'SO:0001060', # sequence variant
+ 'Bio::EnsEMBL::Variation::StructuralVariationFeature' => 'SO:0001537', # structural variant
+ 'Bio::EnsEMBL::Compara::ConstrainedElement' => 'SO:0001009', #DNA_constraint_sequence ????
+ 'Bio::EnsEMBL::Funcgen::RegulatoryFeature' => 'SO:0001679', # transcription_regulatory_region
+);
+
+=head2 new
+
+ Constructor
+ Arg [1] : OntologyAdaptor from the EnsEMBL registry
+ Returntype : Bio::EnsEMBL::BiotypeMapper
+
+=cut
+
+sub new {
+ my $class = shift;
+ my $self = {
+ ontology_adaptor => shift,
+ };
+
+ bless $self, $class;
+ return $self;
+}
+
+=head2 translate_feature_to_SO_term
+
+ Arg [0] : Bio::EnsEMBL::Feature, subclass or related Storable
+ Description: Translates a Feature type into an SO term. If the Feature is a
+ Gene or Transcript, then a further refinement of the type is made
+ via Biotype
+ Returntype : String
+
+=cut
+
+sub translate_feature_to_SO_term {
+ my $self = shift;
+ my $feature = shift;
+ my $so_accession;
+ my $so_term;
+ if (ref($feature) eq "Bio::EnsEMBL::Gene" and exists $gene_so_mapping{$feature->biotype}) {
+ $so_accession = $gene_so_mapping{$feature->biotype};
+ }
+ elsif (ref($feature) eq "Bio::EnsEMBL::Transcription" and exists $transcript_so_mapping{$feature->biotype}) {
+ $so_accession = $transcript_so_mapping{$feature->biotype};
+ }
+ else {
+ $so_accession = $feature_so_mapping{ref($feature)};
+ }
+ if (defined($so_accession)) {
+ $so_term = $self->{'ontology_adaptor'}->fetch_by_accession($so_accession);
+ }
+ else {
+ carp "Ontology mapping not found for ".ref($feature)."\n";
+ return "????????";
+ }
+
+ return $so_term->name;
+}
+
+
+=head2 translate_SO_to_biotype
+
+ Arg [0] : Sequence Ontology term, either in name or URI format
+ Description: Returns the closest corresponding Ensembl biotypes to a given SO term
+ Returntype : String containing a comma-separated list of biotypes
+=cut
+
+sub translate_SO_to_biotype {
+ my $self = shift;
+ my $translate_me = shift;
+
+ my @so_names;
+# look up text in ontology database
+ if ($translate_me !~ /^SO:/) {
+ my $so_terms = $self->{'ontology_adaptor'}->fetch_all_by_name($translate_me);
+ @so_names = [];
+ foreach my $term (@{$so_terms}) {
+ push @so_names,$term->accession();
+ }
+ }
+ else {
+ push @so_names,$translate_me;
+ }
+# convert list of accessions into biotypes
+ my @biotypes;
+ foreach my $accession (@so_names) {
+ foreach my $key (keys %gene_so_mapping) {
+ if ($gene_so_mapping{$key} eq $accession) {
+ push @biotypes,$key;
+ }
+ }
+ foreach my $key (keys %transcript_so_mapping) {
+ if ($transcript_so_mapping{$key} eq $accession) {
+ push @biotypes,$key;
+ }
+ }
+ foreach my $key (keys %feature_so_mapping) {
+ if ($feature_so_mapping{$key} eq $accession) {
+ push @biotypes,$key;
+ }
+ }
+ }
+
+ return join (',',@biotypes);
+}
+
+1;
diff --git a/modules/Bio/EnsEMBL/Utils/IO/GFFSerializer.pm b/modules/Bio/EnsEMBL/Utils/IO/GFFSerializer.pm
new file mode 100644
index 0000000000..345564f64b
--- /dev/null
+++ b/modules/Bio/EnsEMBL/Utils/IO/GFFSerializer.pm
@@ -0,0 +1,214 @@
+=pod
+
+=head1 LICENSE
+
+ Copyright (c) 1999-2011 The European Bioinformatics Institute and
+ Genome Research Limited. All rights reserved.
+
+ This software is distributed under a modified Apache license.
+ For license details, please see
+
+ http://www.ensembl.org/info/about/code_licence.html
+
+=head1 NAME
+
+GFFSerializer - Feature to GFF converter
+
+=head1 AUTHOR
+
+Kieron Taylor, 2011 - ktaylor@ebi.ac.uk
+
+=head1 SYNOPSIS
+
+use Bio::EnsEMBL::Utils::IO::GFFSerializer;
+use Bio::EnsEMBL::Utils::BiotypeMapper;
+
+my $ontology_adaptor = $registry->get_adaptor( 'Multi', 'Ontology', 'OntologyTerm' );
+my $biotype_mapper = new BiotypeMapper($ontology_adaptor);
+my $serializer = new GFFSerializer($biotype_mapper,$output_fh);
+
+my $variation_feature_adaptor = $registry->get_adaptor( $config{'species'}, 'variation', 'variationfeature' );
+$serializer->print_metadata("Variation Features:");
+my $iterator = $variation_feature_adaptor->fetch_Iterator_by_Slice($slice,undef,60000);
+$serializer->print_feature_Iterator($iterator);
+
+=head1 DESCRIPTION
+
+Subclass of Serializer that can turn a feature into a line for the GFF3 format. Requires
+a BiotypeMapper in order to translate biotypes to SO terms.
+
+=cut
+
+package Bio::EnsEMBL::Utils::IO::GFFSerializer;
+use strict;
+use warnings;
+use Bio::EnsEMBL::Utils::Exception;
+use Bio::EnsEMBL::Utils::BiotypeMapper;
+use URI::Escape;
+use Bio::EnsEMBL::Utils::IO::Serializer;
+
+use base qw(Bio::EnsEMBL::Utils::IO::Serializer);
+
+my %strand_conversion = ( '1' => '+', '0' => '?', '-1' => '-');
+
+=head2 new
+
+ Constructor
+ Arg [1] : BiotypeMapper
+ Arg [2] : Optional File handle
+
+ Returntype : Bio::EnsEMBL::Utils::IO::Serializer
+
+=cut
+
+sub new {
+ my $class = shift;
+ my $self = {
+ mapper => shift,
+ filehandle => shift,
+ };
+ bless $self, $class;
+ if ( ref($self->mapper) ne "Bio::EnsEMBL::Utils::BiotypeMapper" ) {
+ throw("GFF format requires an instance of Bio::EnsEMBL::Utils::BiotypeMapper to function");
+ }
+
+ if (!defined ($self->{'filehandle'})) {
+ # no file handle, let the handle point to a copy of STDOUT instead
+ open $self->{'filehandle'}, ">&STDOUT";
+ $self->{'stdout'} = 1;
+ }
+ return $self;
+}
+
+=head2 print_feature
+
+ Arg [1] : Bio::EnsEMBL::Feature, subclass or related pseudo-feature
+ Example : $reporter->print_feature($feature,$slice_start_coordinate,"X")
+ Description: Asks a feature for its summary, and generates a GFF3 compliant entry to hand back again
+ Additional attributes are handed through to column 9 of the output using exact spelling
+ and capitalisation of the feature-supplied hash.
+ Returntype : none
+=cut
+
+sub print_feature {
+ my $self = shift;
+ my $feature = shift;
+ my $biotype_mapper = $self->{'mapper'};
+
+ my $text_buffer = "";
+ if ($feature->can('summary_as_hash') ) {
+ my %summary = %{$feature->summary_as_hash};
+ my $row = "";
+# Column 1 - seqname, the name of the sequence/chromosome the feature is on. Landmark for start below
+ if (!defined($summary{'seq_region_name'})) {$summary{'seq_region_name'} = "?";}
+ $row .= $summary{'seq_region_name'}."\t";
+
+# Column 2 - source, complicated with Ensembl not being the originator of all data
+ $row .= "EnsEMBL\t";
+
+# Column 3 - feature, the ontology term for the kind of feature this row is
+ my $so_term = $biotype_mapper->translate_feature_to_SO_term($feature);
+ $row .= $so_term."\t";
+
+# Column 4 - start, the start coordinate of the feature, here shifted to chromosomal coordinates
+# Start and end must be in ascending order for GFF. Circular genomes require the length of
+# the circuit to be added on.
+ if ($summary{'start'} > $summary{'end'}) {
+ #assumes this is not a Compara circular sequence and can treat is as a Feature
+ if ($feature->slice() && $feature->slice()->is_circular() ) {
+ $summary{'end'} = $summary{'end'} + $feature->seq_region_length;
+ }
+ # non-circular, but end still before start
+ else {$summary{'end'} = $summary{'start'};}
+ }
+ $row .= $summary{'start'} . "\t";
+
+# Column 5 - end, coordinates (absolute) for the end of this feature
+ $row .= $summary{'end'} . "\t";
+
+# Column 6 - score, for variations only.
+ if (exists($summary{'score'})) {
+ $row .= $summary{'score'}."\t";
+ }
+ else {
+ $row .= ".\t";
+ }
+
+# Column 7 - strand, up or down
+ if (exists($summary{'strand'})) {
+ $row .= $strand_conversion{$summary{'strand'}}."\t";
+ }
+ else {
+ $row .= ".\t";
+ }
+
+# Column 8 - reading frame, necessary only for Exons
+ $row .= ".\t";
+
+# Column 9 - the 'other' section for all GFF and GVF compliant attributes
+# We include Stable ID and biotype where possible to supplement the information in the other columns
+ delete $summary{'seq_region_start'};
+ delete $summary{'seq_region_name'};
+ delete $summary{'start'};
+ delete $summary{'end'};
+ delete $summary{'strand'};
+ delete $summary{'score'};
+# Slice the hash for specific keys in GFF-friendly order
+ my @ordered_keys = qw(ID Name Alias Parent Target Gap Derives_from Note Dbxref Ontology_term Is_circular);
+ my @ordered_values = @summary{@ordered_keys};
+ while (my $key = shift @ordered_keys) {
+ my $value = shift @ordered_values;
+ if ($value) {
+ $row .= $key."=".uri_escape($value,'\t\n\r;=%&,').";";
+ }
+ delete $summary{$key};
+ }
+# Catch the remaining keys, containing whatever else the Feature provided
+ foreach my $attribute ( keys(%summary)) {
+ if (ref $summary{$attribute} eq "ARRAY") {
+ $row .= $attribute."=".join (',',@{$summary{$attribute}}) . ";"
+ }
+ else {
+ if ($summary{$attribute}) { $row .= $attribute."=".uri_escape($summary{$attribute},'\t\n\r;=%&,') . ";"; }
+ }
+ }
+# trim off any trailing commas left by the ordered keys stage above:
+ $text_buffer .= $row."\n";
+ }
+ else {
+ warning("Feature failed to self-summarise");
+ }
+ #filehandle is inherited
+ my $fh = $self->{'filehandle'};
+ print $fh $text_buffer;
+}
+
+=head2 print_main_header
+
+ Arg [1] : Arrayref of slices going into the file.
+ Description: Printing the header text or metadata required for GFF,
+ using a list of slices to be written
+ Returntype : None
+=cut
+
+sub print_main_header {
+ my $self = shift;
+ my $arrayref_of_slices = shift;
+ my $fh = $self->{'filehandle'};
+
+ print $fh "##gff-version 3\n";
+ foreach my $slice (@{$arrayref_of_slices}) {
+ if (not defined($slice)) { warning("Slice not defined.\n"); return;}
+ print $fh "##sequence-region ",$slice->seq_region_name," ",$slice->start," ",$slice->end,"\n";
+ }
+}
+
+sub print_metadata {
+ my $self = shift;
+ my $text = shift;
+ my $fh = $self->{'filehandle'};
+ print $fh "\n# ".$text."\n";
+}
+
+
+1;
diff --git a/modules/Bio/EnsEMBL/Utils/IO/ReportSerializer.pm b/modules/Bio/EnsEMBL/Utils/IO/ReportSerializer.pm
new file mode 100644
index 0000000000..ae400328c9
--- /dev/null
+++ b/modules/Bio/EnsEMBL/Utils/IO/ReportSerializer.pm
@@ -0,0 +1,209 @@
+=pod
+
+=head1 LICENSE
+
+ Copyright (c) 1999-2011 The European Bioinformatics Institute and
+ Genome Research Limited. All rights reserved.
+
+ This software is distributed under a modified Apache license.
+ For license details, please see
+
+ http://www.ensembl.org/info/about/code_licence.html
+
+=head1 NAME
+
+Report Serializer - generating textual summary reports
+
+=head1 AUTHOR
+
+Kieron Taylor, 2011 - ktaylor@ebi.ac.uk
+
+=head1 SYNOPSIS
+
+ use Bio::EnsEMBL::Registry;
+ use Bio::EnsEMBL::Utils::IO::ReportSerializer;
+ use IO::File;
+
+ my $registry = 'Bio::EnsEMBL::Registry';
+ $output_fh = IO::File->new($config{'output'},'w') or die;
+ $serializer = new ReportSerializer($output_fh);
+ my $slice_adaptor = $registry->get_adaptor( 'Human', 'Core', 'Slice' );
+ my $slice = $slice_adaptor->fetch_by_toplevel_location("6:1000000..1500000");
+
+ $serializer->print_section_header($slice);
+ $serializer->print_feature_list($slice->get_all_Genes);
+
+=head1 DESCRIPTION
+
+Subclass of Serializer that can turn a feature into a text block
+Unsuited to very large slices, because it requires a select-all approach for features.
+
+=cut
+
+package Bio::EnsEMBL::Utils::IO::ReportSerializer;
+use strict;
+use warnings;
+use Bio::EnsEMBL::Utils::Exception;
+use URI::Escape;
+use Bio::EnsEMBL::Utils::IO::Serializer;
+
+use base qw(Bio::EnsEMBL::Utils::IO::Serializer);
+
+my %strand_conversion = ( '1' => '+', '0' => '?', '-1' => '-');
+
+my %feature_conversion = ( 'Bio::EnsEMBL::Gene' => 'Gene',
+ 'Bio::EnsEMBL::Transcript' => 'Transcript',
+ 'Bio::EnsEMBL::Translation' => 'Translation',
+ 'Bio::EnsEMBL::Variation::StructuralVariationFeature' => 'Structural Variation',
+ 'Bio::EnsEMBL::Variation::VariationFeature' => 'Variation',
+ 'Bio::EnsEMBL::Funcgen::RegulatoryFeature' => 'Regulatory Feature',
+ 'Bio::EnsEMBL::Compara::ConstrainedElement' => 'Constrained Element',
+);
+
+# Hash for selecting the correct attributes of unseen features for crude summary. This hash is
+# for fallback behaviour, slicing summary hashes for a limited set of values.
+my %printables = (
+ 'Bio::EnsEMBL::Gene' => ['ID','biotype','start','end'],
+ 'Bio::EnsEMBL::Transcript' => ['ID','start','end'],
+ 'Bio::EnsEMBL::Translation' => ['ID'],
+ 'Bio::EnsEMBL::Variation::VariationFeature' => ['ID','start','end','strand','seq_region_name'],
+ 'Bio::EnsEMBL::Variation::StructuralVariationFeature' => ['ID','start','end','strand','seq_region_name'],
+ 'Bio::EnsEMBL::Funcgen::RegulatoryFeature' => ['ID','start','end','strand'],
+ 'Bio::EnsEMBL::Compara::ConstrainedElement' => ['ID','start','end','strand','seq_region_name'],
+ );
+
+=head2 print_feature
+
+ Arg [1] : Bio::EnsEMBL::Feature, subclass or related pseudo-feature
+ Example : $reporter->print_feature($feature,$slice_start_coordinate,"X")
+=cut
+
+sub print_feature {
+ my $self = shift;
+ my $feature = shift;
+ my $fh = $self->{'filehandle'};
+ my $feature_type = ref($feature);
+
+ if ($feature->can('summary_as_hash') ) {
+ my %summary = %{$feature->summary_as_hash};
+ my @values = @summary{ @{$printables{$feature_type}} };
+ print $fh join(',',@values)."\n";
+ }
+ else {
+ warning("Feature failed to self-summarise");
+ }
+}
+
+=head2 print_feature_list
+
+ Arg [1] : Listref of Bio::EnsEMBL::Feature, subclass or related pseudo-feature
+ Description: Relies on a list of similar features to print in a block together.
+ Overrides superclass method
+ Results are truncated after the first 100 features for brevity.
+ Example : $reporter->print_feature_list(\@features);
+=cut
+
+sub print_feature_list {
+ my $self = shift;
+ my $feature_list = shift;
+ my $fh = $self->{'filehandle'};
+
+ my $example_feature = $feature_list->[0];
+ my $feature_type = ref($example_feature);
+ my $feature_count = 0;
+ unless (defined $feature_type) {$feature_type = "Feature"};
+ print $fh "There are ",scalar(@$feature_list)," ",$feature_conversion{$feature_type},(scalar(@$feature_list) != 1) ? "s":""," in this region\n";
+ if (scalar(@$feature_list) > 100 ) { print $fh "Too many to display, results truncated to the first 100\n";}
+ print $fh "\n";
+ foreach my $feature (@$feature_list) {
+ $feature_count++;
+ my %attributes = %{$feature->summary_as_hash};
+
+ if ($feature_count == 100) {last;}
+ # Begin the feature-specific formatting code
+ if ($feature_type eq "Bio::EnsEMBL::Gene") {
+ print $fh "\tGene ".$feature_count.": ".$attributes{'external_name'}.",".$attributes{'ID'}."\n";
+ print $fh "\tBiotype: ".$attributes{'biotype'}."\n";
+ print $fh "\tLocation: ".$attributes{'start'}."-".$attributes{'end'}." bp\n\n";
+
+ print $fh "\tTranscripts and proteins\n";
+ foreach my $transcript (@{$feature->get_all_Transcripts}) {
+ my %tr_summary = %{$transcript->summary_as_hash};
+ print $fh "\t\t ".$tr_summary{'ID'};
+ my $translation = $transcript->translation;
+ if (defined $translation) {
+ my %pr_summary = %{$translation->summary_as_hash};
+ print $fh " - ".$pr_summary{'ID'}."\n\n";
+ }
+ else {print $fh " - no protein\n\n";}
+ }
+ print $fh "\n";
+ }
+ elsif ($feature_type eq "Bio::EnsEMBL::Funcgen::RegulatoryFeature") {
+ print $fh "\t".$attributes{'ID'}."\n";
+ }
+ elsif ($feature_type eq "Bio::EnsEMBL::Compara::ConstrainedElement") {
+ print $fh "\t".$attributes{'start'}."-".$attributes{'end'}."\n";
+ }
+ elsif ( $feature_type eq "Bio::EnsEMBL::Variation::StructuralVariationFeature"
+ or $feature_type eq "Bio::EnsEMBL::Variation::VariationFeature") {
+ print $fh "\tID: ".$attributes{'ID'}." Position: ".
+ $attributes{'start'}."-".$attributes{'end'}." on strand ".$attributes{'strand'}." \n";
+ }
+ else {
+ # slice favourite values out unformatted.
+ my @values = @attributes{ @{$printables{$feature_type}} };
+ print $fh $feature_type.join(',',@values)."\n";
+
+ }
+ }
+}
+
+# Just print individuals without awareness of list size and position.
+sub print_feature_iterator {
+ my $self = shift;
+ my $feature_iterator = shift;
+ while ($feature_iterator->has_next) {
+ my $feature = $feature_iterator->next;
+ $self->print_feature($feature);
+ }
+}
+
+=head2 print_main_header
+
+ Arg [1] : Arrayref of slices going into the file.
+ Description: Printing the header text for this report
+ Requires a slice list in order to report how many will be printed
+ Returntype : None
+=cut
+
+sub print_main_header {
+ my $self = shift;
+ my $arrayref_of_slices = shift;
+ my $fh = $self->{'filehandle'};
+
+ my $regions = scalar @{$arrayref_of_slices};
+ print $fh "Report for $regions region";
+ if ($regions > 1) { print $fh "s";}
+ print $fh "\n\n";
+}
+
+
+=head2 print_section_header
+
+ Arg [1] : Bio::EnsEMBL::Slice
+ Description: Prints a summary of the slice
+ Intended to be used prior to print_feature_list()
+ Returntype : None
+
+=cut
+
+sub print_section_header {
+ my $self = shift;
+ my $slice = shift;
+ my $fh = $self->{'filehandle'};
+
+ print $fh " Region: ",$slice->seq_region_name," ",$slice->start,"-",$slice->end," bp\n\n";
+
+}
+
diff --git a/modules/Bio/EnsEMBL/Utils/IO/Serializer.pm b/modules/Bio/EnsEMBL/Utils/IO/Serializer.pm
new file mode 100644
index 0000000000..2dff8efdd9
--- /dev/null
+++ b/modules/Bio/EnsEMBL/Utils/IO/Serializer.pm
@@ -0,0 +1,160 @@
+=pod
+
+=head1 LICENSE
+
+ Copyright (c) 1999-2011 The European Bioinformatics Institute and
+ Genome Research Limited. All rights reserved.
+
+ This software is distributed under a modified Apache license.
+ For license details, please see
+
+ http://www.ensembl.org/info/about/code_licence.html
+
+=head1 NAME
+
+Serializer - An abstract serializer for turning EnsEMBL data into other formats
+
+=head1 AUTHOR
+
+Kieron Taylor, 2011 - ktaylor@ebi.ac.uk
+
+=head1 SYNOPSIS
+
+my $serializer = new Serializer( $filehandle );
+$serializer->print_feature_list( \@list_of_features );
+
+=head1 DESCRIPTION
+
+Subclass this class to create a format-specific serializer.
+Be sure to implement print_feature at the bare minimum
+
+=cut
+
+package Bio::EnsEMBL::Utils::IO::Serializer;
+use strict;
+use warnings;
+use Bio::EnsEMBL::Utils::Exception;
+use Bio::EnsEMBL::Utils::SeqDumper;
+
+
+=head2 new
+
+ Constructor
+ Arg [1] : Optional File handle
+ Returntype : Bio::EnsEMBL::Utils::IO::Serializer
+
+=cut
+
+sub new {
+ my $class = shift;
+ my $self = {
+ 'filehandle' => shift,
+ };
+ bless $self, $class;
+ if (!defined ($self->{'filehandle'})) {
+ # no file handle, let the handle point to a copy of STDOUT instead
+ open $self->{'filehandle'}, ">&STDOUT";
+ $self->{'stdout'} = 1;
+ }
+ return $self;
+}
+
+=head2 DESTROY
+
+ Destructor
+ Description: Restores default state of the STDOUT filehandle as it is a copy
+ and may not flush correctly.
+=cut
+
+sub DESTROY {
+ my $self = shift;
+ if ($self->{'stdout'}) {
+ close $self->{'filehandle'};
+ }
+}
+
+sub print_feature {
+ throw( "print_feature method not implemented.");
+}
+
+=head2 print_feature_list
+
+ Arg [1] : Listref of features
+ Description: Run print_feature on every feature in the list
+
+=cut
+
+sub print_feature_list {
+ my $self = shift;
+ my $feature_list = shift;
+ if (ref($feature_list) eq 'ARRAY') {
+ foreach my $feature (@{$feature_list}) {
+ $self->print_feature($feature);
+ }
+ }
+ else {
+ throw( "print_feature_list requires a listref as argument" );
+ }
+}
+
+=head2 print_feature_Iterator
+
+ Arg [1] : Bio::EnsEMBL::Utils::Iterator
+ Description: Automatically spools through an iterator for convenience
+ Returntype : None
+=cut
+
+sub print_feature_Iterator {
+ my $self = shift;
+ my $iterator = shift;
+ if ($iterator->can('has_next')) {
+ $iterator->each(sub {$self->print_feature($_)});
+ }
+ else {
+ throw("Supplied iterator does not look like Bio::EnsEMBL::Utils::Iterator");
+ }
+}
+
+=head2 print_metadata
+
+ Arg [1] : String
+ Description: Pipes a custom string into the filehandle that the serializer is using
+
+=cut
+
+sub print_metadata {
+ my $self = shift;
+ my $text = shift;
+ my $fh = $self->{'filehandle'};
+ print $fh "\n".$text."\n";
+}
+
+=head2 print_main_header
+
+ Arg [1] : Arrayref of slices going into the file.
+ Description: Printing the header text or metadata required for this file format,
+ Re-implement in the serializer.
+ Returntype : None
+=cut
+
+sub print_main_header {
+ my $self = shift;
+# my $arrayref_of_slices = shift;
+# my $fh = $self->{'filehandle'};
+ warning("No writer for headers in this format. Nothing done" );
+}
+
+=head2 print_sequence
+ Arg [1] : Bio::EnsEMBL::Slice
+ Description: By default, prints a block of FASTA format sequence from the given slice
+=cut
+
+sub print_sequence {
+ my $self = shift;
+ my $slice = shift;
+ print "##FASTA\n";
+ Bio::EnsEMBL::Utils::SeqDumper->dump( $slice, 'FASTA', $self->{'filehandle'});
+}
+
+
+1;
--
GitLab