#
# BioPerl module for Bio::SeqFeature::Collection
#
# Please direct questions and support issues to <bioperl-l@bioperl.org>
#
# Cared for by Jason Stajich <jason@bioperl.org>
#
# Copyright Jason Stajich
#
# You may distribute this module under the same terms as perl itself
# POD documentation - main docs before the code
=head1 NAME
Bio::SeqFeature::Collection - A container class for SeqFeatures
suitable for performing operations such as finding features within a
range, that match a certain feature type, etc.
=head1 SYNOPSIS
use Bio::SeqFeature::Collection;
use Bio::Location::Simple;
use Bio::Tools::GFF;
use Bio::Root::IO;
use File::Spec;
# let's first input some features
my $gffio = Bio::Tools::GFF->new(-file => File::Spec->catfile
("t","data","myco_sites.gff"),
-gff_version => 2);
my @features = ();
# loop over the input stream
while(my $feature = $gffio->next_feature()) {
# do something with feature
push @features, $feature;
}
$gffio->close();
# build the Collection object
my $col = Bio::SeqFeature::Collection->new();
# add these features to the object
my $totaladded = $col->add_features(\@features);
my @subset = $col->features_in_range(-start => 1,
-end => 25000,
-strand => 1,
-contain => 0);
# subset should have 18 entries for this dataset
print "size is ", scalar @subset, "\n";
@subset = $col->features_in_range(-range => Bio::Location::Simple->new
(-start => 70000,
-end => 150000,
-strand => -1),
-contain => 1,
-strandmatch => 'strong');
# subset should have 22 entries for this dataset
print "size is ", scalar @subset, "\n";
print "total number of features in collection is ",
$col->feature_count(),"\n";
=head1 DESCRIPTION
This object will efficiently allow one for query subsets of ranges
within a large collection of sequence features (in fact the objects
just have to be Bio::RangeI compliant). This is done by the creation
of bins which are stored in order in a B-Tree data structure as
provided by the DB_File interface to the Berkeley DB.
This is based on work done by Lincoln for storage in a mysql instance
- this is intended to be an embedded in-memory implementation for
easily querying for subsets of a large range set.
Collections can be made persistent by keeping the indexfile and
passing in the -keep flag like this:
my $collection = Bio::SeqFeature::Collection->new(-keep => 1,
-file => 'col.idx');
$collaction->add_features(\@features);
undef $collection;
# To reuse this collection, next time you initialize a Collection object
# specify the filename and the index will be reused.
$collection = Bio::SeqFeature::Collection->new(-keep => 1,
-file => 'col.idx');
=head1 FEEDBACK
=head2 Mailing Lists
User feedback is an integral part of the evolution of this and other
Bioperl modules. Send your comments and suggestions preferably to
the Bioperl mailing list. Your participation is much appreciated.
bioperl-l@bioperl.org - General discussion
http://bioperl.org/wiki/Mailing_lists - About the mailing lists
=head2 Support
Please direct usage questions or support issues to the mailing list:
I<bioperl-l@bioperl.org>
rather than to the module maintainer directly. Many experienced and
reponsive experts will be able look at the problem and quickly
address it. Please include a thorough description of the problem
with code and data examples if at all possible.
=head2 Reporting Bugs
Report bugs to the Bioperl bug tracking system to help us keep track
of the bugs and their resolution. Bug reports can be submitted via
the web:
https://redmine.open-bio.org/projects/bioperl/
=head1 AUTHOR - Jason Stajich
Email jason@bioperl.org
=head1 CONTRIBUTORS
Using code and strategy developed by Lincoln Stein (lstein@cshl.org)
in Bio::DB::GFF implementation. Credit also to Lincoln for suggesting
using Storable to serialize features rather than my previous implementation
which kept the features in memory.
=head1 APPENDIX
The rest of the documentation details each of the object methods.
Internal methods are usually preceded with a _
=cut
# Let the code begin...
package Bio::SeqFeature::Collection;
use strict;
# Object preamble - inherits from Bio::Root::Root
use Bio::DB::GFF::Util::Binning;
use DB_File;
use Bio::Location::Simple;
use Bio::SeqFeature::Generic;
use Storable qw(freeze thaw);
use base qw(Bio::Root::Root Bio::SeqFeature::CollectionI);
# This may need to get re-optimized for BDB usage as these
# numbers were derived empirically by Lincoln on a mysql srv
# running on his laptop
# this is the largest that any reference sequence can be (100 megabases)
use constant MAX_BIN => 100_000_000;
# this is the smallest bin (1 K)
use constant MIN_BIN => 1_000;
=head2 new
Title : new
Usage : my $obj = Bio::SeqFeature::Collection->new();
Function: Builds a new Bio::SeqFeature::Collection object
Returns : Bio::SeqFeature::Collection
Args :
-minbin minimum value to use for binning
(default is 100,000,000)
-maxbin maximum value to use for binning
(default is 1,000)
-file filename to store/read the
BTREE from rather than an in-memory structure
(default is false and in-memory).
-keep boolean, will not remove index file on
object destruction.
-features Array ref of features to add initially
=cut
sub new {
my($class,@args) = @_;
my $self = $class->SUPER::new(@args);
my ($maxbin,$minbin, $file, $keep,
$features) = $self->_rearrange([qw(MAXBIN MINBIN FILE KEEP
FEATURES)],@args);
defined $maxbin && $self->max_bin($maxbin);
defined $minbin && $self->min_bin($minbin);
defined $features && $self->add_features($features);
$DB_BTREE->{'flags'} = R_DUP ;
$DB_BTREE->{'compare'} = \&_compare;
$self->{'_btreehash'} = {};
if( $file ) {
$self->debug("using file $file");
$self->indexfile($file);
}
$self->keep($keep);
$self->{'_btree'} = tie %{$self->{'_btreehash'}},
'DB_File', $self->indexfile, O_RDWR|O_CREAT, 0640, $DB_BTREE;
$self->{'_btree'} || $self->throw("Unable to tie DB_File handle");
return $self;
}
=head2 add_features
Title : add_features
Usage : $collection->add_features(\@features);
Function:
Returns : number of features added
Args : arrayref of Bio::SeqFeatureI objects to index
=cut
sub add_features{
my ($self,$feats) = @_;
if( ref($feats) !~ /ARRAY/i ) {
$self->warn("Must provide a valid Array reference to add_features");
return 0;
}
my $count = 0;
foreach my $f ( @$feats ) {
if( ! $f || ! ref($f) || ! $f->isa('Bio::RangeI') ) {
$self->warn("Must provide valid Bio::RangeI objects to add_features, skipping object '$f'\n");
next;
}
my $bin = bin($f->start,$f->end,$self->min_bin);
my $serialized = &feature_freeze($f);
$self->{'_btree'}->put($bin,$serialized);
if( $f->isa('Bio::SeqFeature::Generic') ) {
$self->debug( "$bin for ". $f->location->to_FTstring(). " matches ".$#{$self->{'_features'}}. "\n");
}
$count++;
}
return $count;
}
=head2 features_in_range
Title : features_in_range
Usage : my @features = $collection->features_in_range($range)
Function: Retrieves a list of features which were contained or overlap the
the requested range (see Args for way to specify overlap or
only those containe)d
Returns : List of Bio::SeqFeatureI objects
Args : -range => Bio::RangeI object defining range to search,
OR
-start => start,
-end => end,
-strand => strand
-contain => boolean - true if feature must be completely
contained with range
OR false if should include features that simply overlap
the range. Default: true.
-strandmatch => 'strong', ranges must have the same strand
'weak', ranges must have the same
strand or no strand
'ignore', ignore strand information
Default. 'ignore'.
=cut
sub features_in_range{
my $self = shift;
my (@args) = @_;
my ($range, $contain, $strandmatch,$start,$end,$strand);
if( @args == 1 ) {
$range = shift @args;
} else {
($start,$end,$strand,$range,
$contain,$strandmatch) = $self->_rearrange([qw(START END
STRAND
RANGE CONTAIN
STRANDMATCH)],
@args);
$contain = 1 unless defined $contain;
}
$strand = 1 unless defined $strand;
if( $strand !~ /^([\-\+])$/ &&
$strand !~ /^[\-\+]?1$/ ) {
$self->warn("must provide a valid numeric or +/- for strand");
return ();
}
if( defined $1 ) { $strand .= 1; }
if( !defined $start && !defined $end ) {
if( ! defined $range || !ref($range) || ! $range->isa("Bio::RangeI") )
{
$self->warn("Must defined a valid Range for the method feature_in_range");
return ();
}
($start,$end,$strand) = ($range->start,$range->end,$range->strand);
}
my $r = Bio::Location::Simple->new(-start => $start,
-end => $end,
-strand => $strand);
my @features;
my $maxbin = $self->max_bin;
my $minbin = $self->min_bin;
my $tier = $maxbin;
my ($k,$v,@bins) = ("",undef);
while ($tier >= $minbin) {
my ($tier_start,$tier_stop) = (bin_bot($tier,$start),
bin_top($tier,$end));
if( $tier_start == $tier_stop ) {
my @vals = $self->{'_btree'}->get_dup($tier_start);
if( scalar @vals > 0 ) {
push @bins, map { thaw($_) } @vals;
}
} else {
$k = $tier_start;
my @vals;
for( my $rc = $self->{'_btree'}->seq($k,$v,R_CURSOR);
$rc == 0;
$rc = $self->{'_btree'}->seq($k,$v, R_NEXT) ) {
last if( $k > $tier_stop || $k < $tier_start);
push @bins, thaw($v);
}
}
$tier /= 10;
}
my %seen = ();
foreach my $t ( map { ref($_) } @bins) {
next if $seen{$t}++;
eval "require $t";
if( $@ ) {
$self->warn("Trying to thaw a stored feature $t which does not appear in your Perl library. $@");
next;
}
}
$strandmatch = 'ignore' unless defined $strandmatch;
return ( $contain ) ? grep { $r->contains($_,$strandmatch) } @bins :
grep { $r->overlaps($_,$strandmatch)} @bins;
}
=head2 remove_features
Title : remove_features
Usage : $collection->remove_features(\@array)
Function: Removes the requested sequence features (based on features
which have the same location)
Returns : Number of features removed
Args : Arrayref of Bio::RangeI objects
=cut
sub remove_features{
my ($self,$feats) = @_;
if( ref($feats) !~ /ARRAY/i ) {
$self->warn("Must provide a valid Array reference to remove_features");
return 0;
}
my $countprocessed = 0;
foreach my $f ( @$feats ) {
next if ! ref($f) || ! $f->isa('Bio::RangeI');
my $bin = bin($f->start,$f->end,$self->min_bin);
my @vals = $self->{'_btree'}->get_dup($bin);
my $vcount = scalar @vals;
foreach my $v ( @vals ) {
# Once we have uniquely identifiable field
# I think it will work better.
if( $v eq &feature_freeze($f) ) {
$self->{'_btree'}->del_dup($bin,$v);
$vcount--;
$countprocessed++;
}
}
if( $vcount == 0 ) {
$self->{'_btree'}->del($bin);
}
}
$countprocessed;
}
=head2 get_all_features
Title : get_all_features
Usage : my @f = $col->get_all_features()
Function: Return all the features stored in this collection (Could be large)
Returns : Array of Bio::RangeI objects
Args : None
=cut
sub get_all_features{
my ($self) = @_;
my @features;
my ($key,$value);
for (my $status = $self->{'_btree'}->seq($key, $value, R_FIRST) ;
$status == 0 ;
$status = $self->{'_btree'}->seq($key, $value, R_NEXT) )
{ next unless defined $value;
push @features, &thaw($value);
}
if( scalar @features != $self->feature_count() ) {
$self->warn("feature count does not match actual count\n");
}
return @features;
}
=head2 min_bin
Title : min_bin
Usage : my $minbin= $self->min_bin;
Function: Get/Set the minimum value to use for binning
Returns : integer
Args : [optional] minimum bin value
=cut
sub min_bin {
my ($self,$min) = @_;
if( defined $min ) {
$self->{'_min_bin'} = $min;
}
return $self->{'_min_bin'} || MIN_BIN;
}
=head2 max_bin
Title : max_bin
Usage : my $maxbin= $self->max_bin;
Function: Get/Set the maximum value to use for binning
Returns : integer
Args : [optional] maximum bin value
=cut
sub max_bin {
my ($self,$max) = @_;
if( defined $max ) {
$self->{'_max_bin'} = $max;
}
return $self->{'max_bin'} || MAX_BIN;
}
=head2 feature_count
Title : feature_count
Usage : my $c = $col->feature_count()
Function: Retrieve the total number of features in the collection
Returns : integer
Args : none
=cut
sub feature_count {
my $self = shift;
my $count = 0;
for ( keys %{$self->{'_btreehash'}} ) {
my $v = $self->{'_btreehash'}->{$_};
next unless defined $v;
$count++;
}
$count;
}
=head2 indexfile
Title : indexfile
Usage : $obj->indexfile($newval)
Function: Get/set the filename where index is kept
Returns : value of indexfile (a filename string)
Args : on set, new value (a filename string )
=cut
sub indexfile{
my $self = shift;
return $self->{'indexfile'} = shift if @_;
return $self->{'indexfile'};
}
=head2 keep
Title : keep
Usage : $obj->keep($newval)
Function: Get/set boolean flag to keep the indexfile after
exiting program
Example :
Returns : value of keep (boolean)
Args : on set, new value (boolean)
=cut
sub keep{
my $self = shift;
return $self->{'keep'} = shift if @_;
return $self->{'keep'};
}
sub _compare{
if( defined $_[0] && ! defined $_[1]) {
return -1;
} elsif ( defined $_[1] && ! defined $_[0]) {
return 1;
}
$_[0] <=> $_[1];
}
sub feature_freeze {
my $obj = shift;
_remove_cleanup_methods($obj);
return freeze($obj);
}
sub _remove_cleanup_methods {
my $obj = shift;
# we have to remove any cleanup methods here for Storable
for my $funcref ( $obj->_cleanup_methods ) {
$obj->_unregister_for_cleanup($funcref);
}
# ... and the same for any contained features; hopefully any implementations
# adhere to implementing Bio::SeqFeatureI::sub_SeqFeature
for my $contained ($obj->sub_SeqFeature) {
_remove_cleanup_methods($contained);
}
1;
}
sub feature_thaw {
return thaw(shift);
}
sub DESTROY {
my $self = shift;
$self->{'_btree'} = undef;
untie(%{$self->{'_btreehash'}});
if( ! $self->keep && $self->indexfile ) {
my $f = $self->indexfile;
$self->debug( "unlinking ".$f. "\n");
close($f);
unlink($f);
}
$self->SUPER::DESTROY();
}
1;