static module to analyse stable ID mapping results

modules/Bio/EnsEMBL/IdMapping/ResultAnalyser.pm

+package Bio::EnsEMBL::IdMapping::ResultAnalyser;
+
+=head1 NAME
+
+
+=head1 SYNOPSIS
+
+
+=head1 DESCRIPTION
+
+
+=head1 METHODS
+
+
+=head1 LICENCE
+
+This code is distributed under an Apache style licence. Please see
+http:#www.ensembl.org/info/about/code_licence.html for details.
+
+=head1 AUTHOR
+
+Patrick Meidl <meidl@ebi.ac.uk>, Ensembl core API team
+
+=head1 CONTACT
+
+Please post comments/questions to the Ensembl development list
+<ensembl-dev@ebi.ac.uk>
+
+=cut
+
+
+use strict;
+use warnings;
+no warnings 'uninitialized';
+
+use Bio::EnsEMBL::IdMapping::BaseObject;
+our @ISA = qw(Bio::EnsEMBL::IdMapping::BaseObject);
+
+use Bio::EnsEMBL::Utils::Exception qw(throw warning);
+use Bio::EnsEMBL::Utils::ScriptUtils qw(path_append);
+
+
+sub analyse {
+  my $self = shift;
+  my $gene_mappings = shift;
+  my $similarity_events = shift;
+  
+  # argument check
+  unless ($gene_mappings and
+          $gene_mappings->isa('Bio::EnsEMBL::IdMapping::MappingList')) {
+    throw("Need a Bio::EnsEMBL::IdMapping::MappingList of genes.");
+  }
+
+  unless ($similarity_events and ref($similarity_events) eq 'ARRAY') {
+    throw("Need a list of similarity events.");
+  }
+
+  # classify source genes by type (status-logic_name-biotype)
+  $self->classify_source_genes_by_type;
+
+  # classify source genes by mapping status
+  $self->classify_genes_by_mapping($gene_mappings, $similarity_events);
+}
+
+#
+# Analyse stable ID data from two existing dbs.
+# This is for potential stand-alone use of this module.
+#
+# [todo]
+sub analyse_db {
+}
+
+
+sub classify_source_genes_by_type {
+  my $self = shift;
+
+  foreach my $s_gene (values %{ $self->cache->get_by_name('genes_by_id', 'source') }) {
+    $self->add('source', $self->class_key($s_gene), 'all', $s_gene->stable_id);
+  }
+}
+
+
+sub classify_target_genes_by_type {
+  my $self = shift;
+  my $gene_mappings = shift;
+
+  # argument check
+  unless ($gene_mappings and
+          $gene_mappings->isa('Bio::EnsEMBL::IdMapping::MappingList')) {
+    throw("Need a Bio::EnsEMBL::IdMapping::MappingList of genes.");
+  }
+
+  my %result = ();
+  
+  # firrst, create a lookup hash of source genes by target internal ID
+  my %source_genes_by_target = ();
+  foreach my $e (@{ $gene_mappings->get_all_Entries }) {
+    my $s_gene = $self->cache->get_by_key('genes_by_id', 'source', $e->source);
+    my $t_gene = $self->cache->get_by_key('genes_by_id', 'target', $e->target);
+    $source_genes_by_target{$t_gene->id} = $s_gene;
+  }
+
+  # now loop over target genes
+  foreach my $t_gene (values %{ $self->cache->get_by_name('genes_by_id', 'target') }) {
+  
+    # check if target gene has all required properties set
+    unless ($t_gene->status and $t_gene->logic_name and $t_gene->biotype) {
+      $self->logger->warning("Missing data for target gene: ".
+        $t_gene->to_string."\n", 1);
+    }
+
+    my $class = $self->class_key($t_gene);
+
+    # classify as '1' if mapped (using source gene's stable ID), otherwise '0'
+    if (my $s_gene = $source_genes_by_target{$t_gene->id}) {
+      $self->add('target', $class, 'mapped', $s_gene->stable_id);
+    } else {
+      $self->add('target', $class, 'unmapped', $t_gene->stable_id);
+    }
+
+  }
+}
+
+
+#
+# genes will be classified as:
+#   - mapped
+#   - deleted
+#     - lost_similar
+#     - lost_definite
+#
+sub classify_genes_by_mapping {
+  my $self = shift;
+  my $gene_mappings = shift;
+  my $similarity_events = shift;
+  
+  # argument check
+  unless ($gene_mappings and
+          $gene_mappings->isa('Bio::EnsEMBL::IdMapping::MappingList')) {
+    throw("Need a Bio::EnsEMBL::IdMapping::MappingList of genes.");
+  }
+
+  unless ($similarity_events and ref($similarity_events) eq 'ARRAY') {
+    throw("Need a list of similarity events.");
+  }
+
+  # mapped genes
+  foreach my $e (@{ $gene_mappings->get_all_Entries }) {
+    my $s_gene = $self->cache->get_by_key('genes_by_id', 'source', $e->source);
+    $self->add('source', $self->class_key($s_gene), 'mapped',
+      $s_gene->stable_id);
+  }
+
+  # lookup hash for similarities
+  my %similar = ();
+  foreach my $event (@{ $similarity_events }) {
+    my ($stable_id) = split("\t", $event);
+    $similar{$stable_id} = 1;
+  }
+  
+  # deleted genes
+  foreach my $s_gene (values %{ $self->cache->get_by_name('genes_by_id', 'source') }) {
+    
+    my $stable_id = $s_gene->stable_id;
+    my $class = $self->class_key($s_gene);
+
+    unless ($self->get('source', $class, 'mapped', $stable_id)) {
+
+      # sub-classify as 'lost_similar' or 'lost_definite'
+      if ($similar{$stable_id}) {
+        $self->add('source', $class, 'lost_similar', $stable_id);
+      } else {
+        $self->add('source', $class, 'lost_definite', $stable_id);
+      }
+
+    }
+  }
+  
+}
+
+
+#
+# Add a stable ID / property pair to a name/dbtype lookup hash.
+#
+# This datastructure is a bloat for some applications, but is general enough to
+# be used as a lookup hash and to generate statistics (counts by type) and
+# debug lists (dump by type).
+#
+sub add {
+  my ($self, $dbtype, $class, $subclass, $stable_id, $val) = @_;
+
+  # private method, so no argument check done for performance reasons
+
+  # default to a value of '1'
+  $val = 1 unless (defined($val));
+
+  $self->{$dbtype}->{$class}->{$subclass}->{$stable_id} = $val;
+}
+
+
+sub get {
+  my ($self, $dbtype, $class, $subclass, $stable_id) = @_;
+
+  # private method, so no argument check done for performance reasons
+
+  return $self->{$dbtype}->{$class}->{$subclass}->{$stable_id};
+}
+
+
+sub get_all_by_subclass {
+  my ($self, $dbtype, $class, $subclass) = @_;
+
+  # argument check
+  throw("Need a dbtype (source|target).") unless ($dbtype);
+  throw("Need a class.") unless ($class);
+  throw("Need a subclass.") unless ($subclass);
+
+  return [ keys %{ $self->{$dbtype}->{$class}->{$subclass} || {} } ];
+}
+
+
+sub get_all_by_class {
+  my ($self, $dbtype, $class) = @_;
+
+  # argument check
+  throw("Need a dbtype (source|target).") unless ($dbtype);
+  throw("Need a class.") unless ($class);
+
+  my %merged = ();
+
+  foreach my $subclass (keys %{ $self->{$dbtype}->{$class} || {} }) {
+    while (my ($key, $val) = each(%{ $self->{$dbtype}->{$class}->{$subclass} })) {
+      $merged{$key} = $val;
+    }
+  }
+
+  return [ keys %merged ];
+}
+
+
+sub get_count_by_subclass {
+  my ($self, $dbtype, $class, $subclass) = @_;
+
+  # argument check
+  throw("Need a dbtype (source|target).") unless ($dbtype);
+  throw("Need a class.") unless ($class);
+  throw("Need a subclass.") unless ($subclass);
+
+  return scalar(keys %{ $self->{$dbtype}->{$class}->{$subclass} || {} });
+}
+
+
+sub get_count_by_class {
+  my ($self, $dbtype, $class) = @_;
+
+  # argument check
+  throw("Need a dbtype (source|target).") unless ($dbtype);
+  throw("Need a class.") unless ($class);
+
+  return scalar(@{ $self->get_all_by_class($dbtype, $class) });
+}
+
+
+sub get_all_classes {
+  my ($self, $dbtype) = @_;
+
+  # argument check
+  throw("Need a dbtype (source|target).") unless ($dbtype);
+
+  return [ sort keys %{ $self->{$dbtype} || {} } ];
+}
+
+
+sub class_key {
+  my ($self, $gene) = @_;
+  return join('-', map { $gene->$_ } qw(status logic_name biotype));
+}
+
+
+sub write_results_to_file { 
+  my $self = shift;
+
+  my $fh = $self->get_filehandle('gene_detailed_mapping_stats.txt', 'stats');
+
+  my $fmt1 = "%-60s%-16s%-16s%-16s\n";
+  my $fmt2 = "%-60s%5.0f (%7s) %5.0f (%7s) %5.0f (%7s)\n";
+  my $fmt3 = "%3.2f%%";
+
+  print $fh "Gene detailed mapping results:\n\n";
+
+  print $fh sprintf($fmt1, "Gene type", "mapped", "lost (similar)",
+    "lost (definite)");
+
+  print $fh ('-'x108), "\n";
+
+  foreach my $class (@{ $self->get_all_classes('source') }) {
+    next if ($class eq 'all');
+
+    my $total = $self->get_count_by_class('source', $class);
+
+    # avoid division by zero error
+    unless ($total) {
+      $self->logger->warning("No count found for $class.\n", 1);
+      next;
+    }
+    
+    my $mapped = $self->get_count_by_subclass('source', $class, 'mapped');
+    my $similar = $self->get_count_by_subclass('source', $class,
+      'lost_similar');
+    my $lost = $self->get_count_by_subclass('source', $class, 'lost_definite');
+
+    print $fh sprintf($fmt2,
+                      $class,
+                      $mapped,  sprintf($fmt3, $mapped/$total*100),
+                      $similar, sprintf($fmt3, $similar/$total*100),
+                      $lost,    sprintf($fmt3, $lost/$total*100));
+  }
+
+  close($fh);
+}
+
+
+sub create_clicklist {
+  my $self = shift;
+
+  my $fh = $self->get_filehandle('genes_lost.html', 'stats');
+
+  # start html output
+  print $fh qq(<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">\n);
+  print $fh qq(<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en-gb"  lang="en-gb">);
+  print $fh "<head>\n";
+  print $fh "<title>Lost genes ";
+  print $fh $self->conf->param('sourcedbname'), ' -&gt; ',
+            $self->conf->param('targetdbname');
+  print $fh "</title>\n";
+  print $fh "</head>\n<body>\n";
+
+  my $prefix = $self->conf->param('urlprefix');
+  unless ($prefix) {
+    $self->logger->warning("No urlprefix set, clicklists might not be useable.\n", 1);
+  }
+
+  my $navigation;
+  my $clicklist;
+
+  foreach my $class (@{ $self->get_all_classes('source') }) {
+    next if ($class eq 'all');
+
+    $navigation .= "$class ";
+    $clicklist .= "<h1>$class</h1>\n";
+    
+    foreach my $subclass (qw(lost_similar lost_definite)) {
+
+      # navigation
+      $navigation .= qq(<a href="#${class}-$subclass">$subclass</a> );
+      
+      # clicklist
+      $clicklist .= "<h2>$subclass</h2>\n";
+
+      foreach my $stable_id (@{ $self->get_all_by_subclass('source', $class, $subclass) }) {
+        $clicklist .= qq(<a href="${prefix}$stable_id">$stable_id</a><br />\n);
+      }
+      
+    }
+
+    $navigation .= "<br />\n";
+  }
+
+  # print navigation and clicklist
+  print $fh "$navigation\n\n";
+  print $fh "$clicklist\n\n";
+
+  # html footer
+  print $fh "</body></html>\n";
+
+  close($fh);
+}
+
+
+sub create_summary_email {
+  my $self = shift;
+  
+  my $fh = $self->get_filehandle('summary_email.txt');
+
+  #
+  # title
+  # 
+  print $fh qq(Stable ID mapping results\n);
+  print $fh qq(=========================\n\n);
+
+  #
+  # parameters used for this run
+  #
+  print $fh $self->conf->list_param_values;
+  print $fh "\n";
+
+  #
+  # mapping stats
+  #
+  foreach my $type (qw(exon transcript translation gene_detailed)) {
+    my $filename = "${type}_mapping_stats.txt";
+    
+    if ($self->file_exists($filename, 'stats')) {
+      print $fh $self->read_from_file($filename, 'stats');
+      print $fh "\n\n";
+    } else {
+      print $fh "No mapping stats found for $type.\n\n";
+    }
+  }
+
+  #
+  # db uploads
+  #
+  my @uploads = (
+    ['stable_ids'  => 'Stable IDs'],
+    ['events'      => 'Stable ID events and mapping session'],
+    ['archive'     => 'Gene and peptide archive'],
+  );
+  
+  my $fmt1 = "%-40s%-20s\n";
+
+  print $fh qq(Data uploaded to db:\n);
+  print $fh qq(====================\n\n);
+
+  if ($self->conf->param('dry_run')) {
+   
+    print $fh "None (dry run).\n";
+  
+  } else {
+  
+    foreach my $u (@uploads) {
+      my $uploaded = 'no';
+      $uploaded = 'yes' if ($self->conf->is_true("upload_".$u->[0]));
+      print $fh sprintf($fmt1, $u->[1], $uploaded);
+    }
+    
+  }
+
+  print $fh "\n";
+
+  #
+  # stats and clicklist
+  #
+  my @output = (
+    ['stats'    => 'statistics (including clicklists of deleted IDs)'],
+    ['debug'    => 'detailed mapping output for debugging'],
+    ['tables'   => 'data files for db upload'],
+  );
+  
+  my $fmt2 = "%-20s%-50s\n";
+
+  print $fh qq(\nOutput directories:\n);
+  print $fh qq(===================\n\n);
+
+  print $fh sprintf($fmt2, qw(DIRECTORY DESCRIPTION));
+  print $fh ('-'x72), "\n";
+
+  print $fh sprintf($fmt2, 'basedir', $self->conf->param('dumppath'));
+
+  foreach my $o (@output) {
+    print $fh sprintf($fmt2, '$basedir/'.$o->[0], $o->[1]);
+  }
+
+  print $fh "\n";
+
+  #
+  # clicklist of first 10 deleted genes
+  #
+  print $fh qq(\nFirst 10 deleted known genes:\n);
+  print $fh qq(=============================\n\n);
+
+  my $in_fh = $self->get_filehandle('genes_lost.txt', 'debug', '<');
+  my $prefix = $self->conf->param('urlprefix');
+  my $i;
+  
+  while (<$in_fh>) {
+    last if (++$i > 10);
+    
+    chomp;
+    my ($stable_id, $type) = split(/\s+/);
+    
+    next unless ($type eq 'known');
+
+    print $fh sprintf($fmt2, $stable_id, "${prefix}$stable_id");
+  }
+
+  close($in_fh);
+  
+  print $fh "\n";
+
+  #
+  # timing
+  #
+  print $fh "\nRuntime:\n";
+  print $fh "========\n\n";
+  print $fh $self->logger->runtime, "\n\n";
+
+  close($fh);
+}
+
+
+sub read_from_file {
+  my $self = shift;
+  my $filename = shift;
+  my $append = shift;
+
+  my $in_fh = $self->get_filehandle($filename, $append, '<');
+
+  my $txt;
+
+  while (<$in_fh>) {
+    $txt .= $_;
+  }
+
+  return $txt;
+}
+
+1;
+