Bio::EnsEMBL::Utils
AssemblyProjector
Toolbar
Summary
Bio::EnsEMBL::Utils::AssemblyProjector -
utility class to post-process projections from one assembly to another
Package variables
No package variables defined.
Included modules
Synopsis
# connect to an old database
my $dba_old = new Bio::EnsEMBL::DBSQL::DBAdaptor(
-host => 'ensembldb.ensembl.org',
-port => 3306,
-user => ensro,
-dbname => 'mus_musculus_core_46_36g',
-group => 'core_old',
);
# connect to the new database containing the mapping between old and
# new assembly
my $dba_new = new Bio::EnsEMBL::DBSQL::DBAdaptor(
-host => 'ensembldb.ensembl.org',
-port => 3306,
-user => ensro,
-dbname => 'mus_musculus_core_47_37',
-group => 'core_new',
);
my $assembly_projector = Bio::EnsEMBL::Utils::AssemblyProjector->new(
-OLD_ASSEMBLY => 'NCBIM36',
-NEW_ASSEMBLY => 'NCBIM37',
-ADAPTOR => $dba_new,
-EXTERNAL_SOURCE => 1,
-MERGE_FRAGMENTS => 1,
-CHECK_LENGTH => 0,
);
# fetch a slice on the old assembly
my $slice_adaptor = $dba_old->get_SliceAdaptor;
my $slice =
$slice_adaptor->fetch_by_region( 'chromosome', 1, undef, undef,
undef, 'NCBIM36' );
my $new_slice = $assembly_projector->old_to_new($slice);
print $new_slice->name, " (", $assembly_projector->last_status, ")\n";
Description
This class implements some utility functions for converting coordinates
between assemblies. A mapping between the two assemblies has to present
the database for this to work, see the 'Related Modules' section below
on how to generate the mapping.
In addition to the "raw" projecting of features and slices, the methods
in this module also apply some sensible rules to the results of the
projection (like discarding unwanted results or merging fragmented
projections). These are the rules (depending on configuration):
Discard the projected feature/slice if:
1. it doesn't project at all (no segments returned)
2. [unless MERGE_FRAGMENTS is set] the projection is fragmented (more
than one segment)
3. [if CHECK_LENGTH is set] the projection doesn't have the same
length as the original feature/slice
4. all segments are on same chromosome and strand
If a projection fails any of these rules, undef is returned instead of
a projected feature/slice. You can use the last_status() method to find
out about the results of the rules tests.
Also note that when projecting features, only a shallow projection is
performed, i.e. other features attached to your features (e.g. the
transcripts of a gene) are not projected automatically, so it will be
the responsability of the user code project all levels of features
involved.
Methods
Methods description
Arg [ADAPTOR] : Bio::EnsEMBL::DBSQL::DBAdaptor $adaptor - a db adaptor for a database containing the assembly mapping Arg [EXTERNAL_SOURCE] : (optional) Boolean $external_source - indicates if source is from a different database Arg [OLD_ASSEMBLY] : name of the old assembly Arg [OLD_ASSEMBLY] : name of the new assembly Arg [OBJECT_TYPE] : (optional) object type ('slice' or 'feature') Arg [MERGE_FRAGMENTS] : (optional) Boolean - determines if segments are merged to return a single object spanning all segments (default: true) Arg [CHECK_LENGTH] : (optional) Boolean - determines if projected objects have to have same length as original (default: false) Example : my $ap = Bio::EnsEMBL::Utils::AssemblyProjector->new( -DBADAPTOR => $dba, -OLD_ASSEMBLY => NCBIM36, -NEW_ASSEMBLY => NCBIM37, ); Description : Constructor. Return type : a Bio::EnsEMBL::Utils::AssemblyProjector object Exceptions : thrown on missing arguments thrown on invalid OBJECT_TYPE Caller : general Status : At Risk : under development |
Arg[1] : Bio::EnsEMBL::Slice or Bio::EnsEMBL::Feature $object - the object to project Example : my $old_slice = $assembly_projector->new_to_old($new_slice, 1); Description : Projects a Slice or Feature from new to old assembly. This method is just a convenience wrapper for $self->project. Return type : same a Arg 1, or undef Exceptions : none Caller : general Status : At Risk : under development |
Arg[1] : Bio::EnsEMBL::Slice or Bio::EnsEMBL::Feature $object - the object to project Example : my $new_slice = $assembly_projector->old_to_new($old_slice); Description : Projects a Slice or Feature from old to new assembly. This method is just a convenience wrapper for $self->project. Return type : same a Arg 1, or undef Exceptions : none Caller : general Status : At Risk : under development |
Arg[1] : Bio::EnsEMBL::Slice or Bio::EnsEMBL::Feature $object - the object to project Arg[2] : String $to_assembly - assembly to project to Example : my $new_slice = $assembly_projector->project($old_slice, 'NCBIM37'); Description : Projects a Slice or Feature to the specified assembly.
Several tests are performed on the result to discard unwanted
results. All projection segments have to be on the same
seq_region and strand. If -MERGE_FRAGMENTS is set, gaps will be
bridged by creating a single object from first_segment_start to
last_segment_end. If -CHECK_LENGTH is set, the projected object
will have to have the same length as the original. You can use
the last_status() method to find out what the result of some of
these rule tests were. Please see the comments in the code for
more details about these rules.
The return value of this method will always be a single object,
or undef if the projection fails any of the rules.
Note that when projecting features, only a "shallow" projection
is performed, i.e. attached features aren't projected
automatically! (e.g. if you project a gene, its transcripts will
have to be projected manually before storing the new gene)
Return type : same a Arg 1, or undef if projection fails any of the rules
Exceptions : thrown on invalid arguments
Caller : general, $self->old_to_new, $self->new_to_old
Status : At Risk
: under development |
Methods code
sub adaptor
{ my $self = shift;
$self->{'adaptor'} = shift if (@_);
return $self->{'adaptor'}; } |
sub check_length
{ my $self = shift;
$self->{'check_length'} = shift if (@_);
return $self->{'check_length'}; } |
sub external_source
{ my $self = shift;
$self->{'external_source'} = shift if (@_);
return $self->{'external_source'}; } |
sub last_status
{ my $self = shift;
$self->{'last_status'} = shift if (@_);
return $self->{'last_status'};
}
1; } |
sub merge_fragments
{ my $self = shift;
$self->{'merge_fragments'} = shift if (@_);
return $self->{'merge_fragments'}; } |
sub new
{ my $caller = shift;
my $class = ref($caller) || $caller;
my ($adaptor, $external_source, $old_assembly, $new_assembly,
$merge_fragments, $check_length) = rearrange([qw(ADAPTOR EXTERNAL_SOURCE
OLD_ASSEMBLY NEW_ASSEMBLY MERGE_FRAGMENTS CHECK_LENGTH)], @_);
unless ($adaptor and ref($adaptor) and
$adaptor->isa('Bio::EnsEMBL::DBSQL::DBAdaptor')) {
throw("You must provide a DBAdaptor to a database containing the assembly mapping.");
}
unless ($old_assembly and $new_assembly) {
throw("You must provide an old and new assembly name.");
}
my $self = {};
bless ($self, $class);
$self->{'adaptor'} = $adaptor;
$self->{'old_assembly'} = $old_assembly;
$self->{'new_assembly'} = $new_assembly;
$self->{'merge_fragments'} = $merge_fragments || 1;
$self->{'check_length'} = $check_length || 0;
$self->{'external_source'} = $external_source || 0;
return $self; } |
sub new_assembly
{ my $self = shift;
$self->{'new_assembly'} = shift if (@_);
return $self->{'new_assembly'}; } |
sub new_to_old
{ my ($self, $object) = @_;
return $self->project($object, $self->old_assembly);
}
} |
sub old_assembly
{ my $self = shift;
$self->{'old_assembly'} = shift if (@_);
return $self->{'old_assembly'}; } |
sub old_to_new
{ my ($self, $object) = @_;
return $self->project($object, $self->new_assembly); } |
sub project
{ my ($self, $object, $to_assembly) = @_;
throw("Need an assembly version to project to.") unless ($to_assembly);
throw("Need an object to project.") unless ($object and ref($object));
my ($slice, $object_type);
if ($object->isa('Bio::EnsEMBL::Feature')) {
$object_type = 'feature';
} elsif ($object->isa('Bio::EnsEMBL::Slice')) {
$object_type = 'slice';
} else {
throw("Need a Feature or Slice to project.");
}
if ($self->external_source) {
my $slice_adaptor = $self->adaptor->get_SliceAdaptor;
if ($object_type eq 'feature') {
my $f_slice = $object->slice;
my $target_slice = $slice_adaptor->fetch_by_name($f_slice->name);
$object->slice($target_slice);
} else {
$object = $slice_adaptor->fetch_by_name($object->name);
}
}
if ($object_type eq 'feature') {
$slice = $object->feature_Slice;
} else {
$slice = $object;
}
if ($slice->coord_system->version eq $to_assembly) {
warning("Assembly version to project to ($to_assembly) is the same as your object's assembly (".$slice->coord_system->version.").");
}
my $cs_name = $slice->coord_system_name;
my @segments = @{ $slice->project($cs_name, $to_assembly) };
if ($slice->strand == -1) {
@segments = reverse(@segments);
}
my @status = ();
return undef unless (@segments);
return undef if (!($self->merge_fragments) and scalar(@segments) > 1);
push @status, 'fragmented' if (scalar(@segments) > 1);
my $first_slice = $segments[0]->to_Slice;
my $last_slice = $segments[-1]->to_Slice;
my $length_mismatch = (($last_slice->end - $first_slice->start + 1) !=
$object->length);
return undef if ($self->check_length and $length_mismatch);
push @status, 'length_mismatch' if ($length_mismatch);
my %sr_names = ();
my %strands = ();
foreach my $seg (@segments) {
my $sl = $seg->to_Slice;
$sr_names{$sl->seq_region_name}++;
$strands{$sl->strand}++;
}
return undef if (scalar(keys %sr_names) > 1 or scalar(keys %strands) > 1);
$self->last_status(join('|', @status));
my $new_slice = $first_slice;
$new_slice->{'end'} = $last_slice->end;
if ($object_type eq 'slice') {
return $new_slice;
} else {
$object->start($new_slice->start);
$object->end($new_slice->end);
$object->strand($new_slice->strand);
$object->slice($new_slice->seq_region_Slice);
$object->dbID(undef);
$object->adaptor(undef);
return $object;
} } |
General documentation
Copyright (c) 1999-2009 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
/info/about/code_licence.html
The process of creating a whole genome alignment between two assemblies
(which is the basis for the use of the methods in this class) is done by
a series of scripts. Please see
ensembl/misc-scripts/assembly/README
for a high-level description of this process, and POD in the individual
scripts for the details.