BaseFeatureAdaptor.pm 31.4 KB
Newer Older
1
#
Simon Potter's avatar
Simon Potter committed
2
# EnsEMBL module for Bio::EnsEMBL::DBSQL::BaseFeatureAdaptor
3
#
Graham McVicker's avatar
Graham McVicker committed
4
# Copyright (c) 2003 Ensembl
5 6 7 8 9
#
# You may distribute this module under the same terms as perl itself

=head1 NAME

10
Bio::EnsEMBL::DBSQL::BaseFeatureAdaptor - An Abstract Base class for all
Graham McVicker's avatar
Graham McVicker committed
11
FeatureAdaptors
12 13 14

=head1 SYNOPSIS

15
Abstract class - should not be instantiated.  Implementation of
16 17 18 19 20
abstract methods must be performed by subclasses.

=head1 DESCRIPTION

This is a base adaptor for feature adaptors. This base class is simply a way
21
of eliminating code duplication through the implementation of methods
22 23
common to all feature adaptors.

24 25
=head1 CONTACT

Graham McVicker's avatar
Graham McVicker committed
26
Contact Ensembl development list for info: <ensembl-dev@ebi.ac.uk>
27 28 29

=head1 METHODS

30 31 32
=cut

package Bio::EnsEMBL::DBSQL::BaseFeatureAdaptor;
33
use vars qw(@ISA);
34 35 36
use strict;

use Bio::EnsEMBL::DBSQL::BaseAdaptor;
37
use Bio::EnsEMBL::Utils::Cache;
38
use Bio::EnsEMBL::Utils::Exception qw(warning throw deprecate stack_trace_dump);
39
use Bio::EnsEMBL::Utils::Argument qw(rearrange);
40 41 42

@ISA = qw(Bio::EnsEMBL::DBSQL::BaseAdaptor);

43 44
our $SLICE_FEATURE_CACHE_SIZE = 4;
our $MAX_SPLIT_QUERY_SEQ_REGIONS = 3;
45 46 47 48 49 50 51 52 53 54 55 56 57 58

=head2 new

  Arg [1]    : list of args @args
               Superclass constructor arguments
  Example    : none
  Description: Constructor which just initializes internal cache structures
  Returntype : Bio::EnsEMBL::BaseFeatureAdaptor
  Exceptions : none
  Caller     : implementing subclass constructors

=cut

sub new {
59
  my $caller = shift;
60

61
  my $class = ref($caller) || $caller;
62

63
  my $self = $class->SUPER::new(@_);
64

65 66 67 68
  #initialize an LRU cache
  my %cache;
  tie(%cache, 'Bio::EnsEMBL::Utils::Cache', $SLICE_FEATURE_CACHE_SIZE);
  $self->{'_slice_feature_cache'} = \%cache;
69 70 71

  return $self;
}
72

73

74

75
# _straight_join
76

77 78 79 80 81 82 83 84 85
#   Arg [1]    : (optional) boolean $new_val
#   Example    : $self->_straight_join(1);
#                $self->generic_fetch($constraint);
#                $self->_straight_join(0);
#   Description: PROTECTED Getter/Setter that turns on/off the use of 
#                a straight join in queries.
#   Returntype : boolean
#   Exceptions : none
#   Caller     : general
86 87 88 89 90 91 92 93 94 95 96

sub _straight_join {
  my $self = shift;
  if(@_) {
    $self->{'_straight_join'} = shift;
  }

  return $self->{'_straight_join'};
}


97 98 99 100
=head2 generic_fetch

  Arg [1]    : (optional) string $constraint
               An SQL query constraint (i.e. part of the WHERE clause)
101 102 103 104 105
  Arg [2]    : (optional) Bio::EnsEMBL::AssemblyMapper $mapper
               A mapper object used to remap features
               as they are retrieved from the database
  Arg [3]    : (optional) Bio::EnsEMBL::Slice $slice
               A slice that features should be remapped to
106
  Example    : $fts = $a->generic_fetch('contig_id in (1234, 1235)', 'Swall');
107 108
  Description: Performs a database fetch and returns feature objects in
               contig coordinates.
109
  Returntype : listref of Bio::EnsEMBL::SeqFeature in contig coordinates
110 111 112 113
  Exceptions : none
  Caller     : BaseFeatureAdaptor, ProxyDnaAlignFeatureAdaptor::generic_fetch

=cut
114

115
sub generic_fetch {
116
  my ($self, $constraint, $mapper, $slice) = @_;
117

118
  my @tabs = $self->_tables;
119
  my $columns = join(', ', $self->_columns());
120

121
  my $db = $self->db();
122 123 124 125 126

  #
  # Construct a left join statement if one was defined, and remove the
  # left-joined table from the table list
  #
127
  my @left_join_list = $self->_left_join();
128 129
  my $left_join = '';
  my @tables;
130 131
  if(@left_join_list) {
    my %left_join_hash = map { $_->[0] => $_->[1] } @left_join_list;
132
    while(my $t = shift @tabs) {
133 134 135
      if( exists $left_join_hash{ $t->[0] } ) {
        my $condition = $left_join_hash{ $t->[0] };
        my $syn = $t->[1];
Web Admin's avatar
Web Admin committed
136
        $left_join .=  "LEFT JOIN\n       ".$t->[0]." $syn ON $condition ";
137
      } else {
138
        push @tables, $t;
139 140 141 142 143
      }
    }
  } else {
    @tables = @tabs;
  }
144

145 146 147 148 149 150 151

  my $straight_join = '';

  if($self->_straight_join()) {
    $straight_join = "STRAIGHT_JOIN";
  }

152 153 154
  #construct a nice table string like 'table1 t1, table2 t2'
  my $tablenames = join(', ', map({ join(' ', @$_) } @tables));

155
  my $sql = "SELECT $straight_join $columns\n  FROM $tablenames $left_join";
156 157 158 159 160

  my $default_where = $self->_default_where_clause;
  my $final_clause = $self->_final_clause;

  #append a where clause if it was defined
161
  if($constraint) {
Web Admin's avatar
Web Admin committed
162
    $sql .= "\n WHERE $constraint ";
163
    if($default_where) {
Web Admin's avatar
Web Admin committed
164
      $sql .= " AND\n       $default_where ";
165 166
    }
  } elsif($default_where) {
Web Admin's avatar
Web Admin committed
167
    $sql .= "\n WHERE $default_where ";
168 169 170
  }

  #append additional clauses which may have been defined
Web Admin's avatar
Web Admin committed
171
  $sql .= "\n$final_clause";
172

173
  my $sth = $db->dbc->prepare($sql);
Web Admin's avatar
Web Admin committed
174
     $sth->execute;
175 176 177

  my $res = $self->_objs_from_sth($sth, $mapper, $slice);

178

179
  return $res;
180 181 182 183
}


=head2 fetch_by_dbID
184 185

  Arg [1]    : int $id
186 187 188
               The unique database identifier for the feature to be obtained
  Example    : $feat = $adaptor->fetch_by_dbID(1234));
               $feat = $feat->transform('contig');
189
  Description: Returns the feature created from the database defined by the
190 191 192 193 194 195 196 197 198
               the id $id.  The feature will be returned in its native
               coordinate system.  That is, the coordinate system in which it
               is stored in the database.  In order to convert it to a
               particular coordinate system use the transfer() or transform()
               method.  If the feature is not found in the database then
               undef is returned instead
  Returntype : Bio::EnsEMBL::Feature or undef
  Exceptions : thrown if $id arg is not provided
               does not exist
199
  Caller     : general
200 201 202 203 204 205

=cut

sub fetch_by_dbID{
  my ($self,$id) = @_;

206
  throw("id argument is required") if(!defined $id);
207

208 209
  #construct a constraint like 't1.table1_id = 123'
  my @tabs = $self->_tables;
210 211
  my ($name, $syn) = @{$tabs[0]};
  my $constraint = "${syn}.${name}_id = $id";
212

213 214 215 216 217
  #Should only be one
  my ($feat) = @{$self->generic_fetch($constraint)};

  return undef if(!$feat);

218
  return $feat;
219 220 221
}


222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237
=head2 fetch_all_by_dbID_list

  Arg [1]    : listref of ints $id_list
               The unique database identifiers for the features to be obtained
  Example    : @feats = @{$adaptor->fetch_by_dbID_list([1234, 2131, 982]))};
  Description: Returns the features created from the database defined by the
               the ids in contained in the id list $id_list.  The features 
               will be returned in their native coordinate system. That is, 
               the coordinate system in which they are stored in the database.
               In order to convert the features to a particular coordinate 
               system use the transfer() or transform() method.  If none of the
               features are found in the database a reference to an empty 
               list is returned.
  Returntype : listref of Bio::EnsEMBL::Features
  Exceptions : thrown if $id arg is not provided
               does not exist
238
  Caller     : general
239 240 241

=cut

242 243
sub fetch_all_by_dbID_list {
  my ($self,$id_list) = @_;
244

245
  if(!defined($id_list) || ref($id_list) ne 'ARRAY') {
246
    throw("id_list list reference argument is required");
247 248
  }

249
  return [] if(!@$id_list);
250

251 252
  my @out;
  #construct a constraint like 't1.table1_id = 123'
253
  my @tabs = $self->_tables;
254
  my ($name, $syn) = @{$tabs[0]};
255

256 257
  # mysql is faster and we ensure that we do not exceed the max query size by
  # splitting large queries into smaller queries of 200 ids
258
  my $max_size = 200;
259

260 261 262 263 264 265 266
  while(@$id_list) {
    my @ids;
    if(@$id_list > $max_size) {
      @ids = splice(@$id_list, 0, $max_size);
    } else {
      @ids = splice(@$id_list, 0);
    }
267

268 269 270 271 272 273
    my $id_str;
    if(@ids > 1)  {
      $id_str = " IN (" . join(',', @ids). ")";
    } else {
      $id_str = " = " . $ids[0];
    }
274

275
    my $constraint = "${syn}.${name}_id $id_str";
276

277 278
    push @out, @{$self->generic_fetch($constraint)};
  }
Arne Stabenau's avatar
Arne Stabenau committed
279

280
  return \@out;
281 282 283 284
}



285
=head2 fetch_all_by_Slice
286 287 288 289 290

  Arg [1]    : Bio::EnsEMBL::Slice $slice
               the slice from which to obtain features
  Arg [2]    : (optional) string $logic_name
               the logic name of the type of features to obtain
291 292 293 294 295 296
  Example    : $fts = $a->fetch_all_by_Slice($slice, 'Swall');
  Description: Returns a listref of features created from the database 
               which are on the Slice defined by $slice. If $logic_name is 
               defined only features with an analysis of type $logic_name 
               will be returned. 
  Returntype : listref of Bio::EnsEMBL::SeqFeatures in Slice coordinates
297 298 299 300 301
  Exceptions : none
  Caller     : Bio::EnsEMBL::Slice

=cut

302
sub fetch_all_by_Slice {
303
  my ($self, $slice, $logic_name) = @_;
304

305
  #fetch by constraint with empty constraint
306
  return $self->fetch_all_by_Slice_constraint($slice, '', $logic_name);
307 308 309
}


310
=head2 fetch_all_by_Slice_and_score
311 312 313

  Arg [1]    : Bio::EnsEMBL::Slice $slice
               the slice from which to obtain features
314
  Arg [2]    : (optional) float $score
315 316 317
               lower bound of the the score of the features retrieved
  Arg [3]    : (optional) string $logic_name
               the logic name of the type of features to obtain
318
  Example    : $fts = $a->fetch_all_by_Slice($slice, 'Swall');
319 320 321 322 323
  Description: Returns a list of features created from the database which are 
               are on the Slice defined by $slice and which have a score 
               greated than $score. If $logic_name is defined, 
               only features with an analysis of type $logic_name will be 
               returned. 
324
  Returntype : listref of Bio::EnsEMBL::SeqFeatures in Slice coordinates
325 326 327 328 329
  Exceptions : none
  Caller     : Bio::EnsEMBL::Slice

=cut

330
sub fetch_all_by_Slice_and_score {
331 332 333
  my ($self, $slice, $score, $logic_name) = @_;
  my $constraint;

334
  if(defined $score) {
335 336 337 338
    #get the synonym of the primary_table
    my @tabs = $self->_tables;
    my $syn = $tabs[0]->[1];
    $constraint = "${syn}.score > $score";
339
  }
340 341
  return $self->fetch_all_by_Slice_constraint($slice, $constraint, 
					      $logic_name);
342
}
343 344


345
=head2 fetch_all_by_Slice_constraint
346

347 348 349 350 351
  Arg [1]    : Bio::EnsEMBL::Slice $slice
               the slice from which to obtain features
  Arg [2]    : (optional) string $constraint
               An SQL query constraint (i.e. part of the WHERE clause)
  Arg [3]    : (optional) string $logic_name
352
               the logic name of the type of features to obtain
353 354
  Example    : $fs = $a->fetch_all_by_Slice_constraint($slc, 'perc_ident > 5');
  Description: Returns a listref of features created from the database which 
355 356 357 358
               are on the Slice defined by $slice and fulfill the SQL 
               constraint defined by $constraint. If logic name is defined, 
               only features with an analysis of type $logic_name will be 
               returned. 
359
  Returntype : listref of Bio::EnsEMBL::SeqFeatures in Slice coordinates
360 361
  Exceptions : thrown if $slice is not defined
  Caller     : Bio::EnsEMBL::Slice
362 363 364

=cut

365
sub fetch_all_by_Slice_constraint {
366 367 368
  my($self, $slice, $constraint, $logic_name) = @_;

  my @result;
369

370
  if(!ref($slice) || !$slice->isa("Bio::EnsEMBL::Slice")) {
Graham McVicker's avatar
Graham McVicker committed
371
    throw("Bio::EnsEMBL::Slice argument expected.");
372 373
  }

374 375
  $constraint ||= '';
  $constraint = $self->_logic_name_to_constraint($constraint, $logic_name);
376 377

  #if the logic name was invalid, undef was returned
378
  return [] if(!defined($constraint));
379

380
  #check the cache and return if we have already done this query
381
  my $key = uc(join(':', $slice->name, $constraint));
382 383 384 385 386

  if(exists($self->{'_slice_feature_cache'}->{$key})) {
    return $self->{'_slice_feature_cache'}->{$key};
  }

387
  my $sa = $slice->adaptor();
388

389 390
  # Hap/PAR support: retrieve normalized 'non-symlinked' slices
  my @proj = @{$sa->fetch_normalized_slice_projection($slice)};
391

392
  if(@proj == 0) {
393 394 395 396
    throw('Could not retrieve normalized Slices. Database contains ' .
          'incorrect assembly_exception information.');
  }

397 398
  # Want to get features on the FULL original slice
  # as well as any symlinked slices
399

400 401
  # Filter out partial slices from projection that are on
  # same seq_region as original slice
402

403
  my $sr_id = $slice->get_seq_region_id();
404

405
  @proj = grep { $_->to_Slice->get_seq_region_id() != $sr_id } @proj;
406

407 408 409
  my $segment = bless([1,$slice->length(),$slice ],
                      'Bio::EnsEMBL::ProjectionSegment');
  push( @proj, $segment );
410

411 412 413 414 415 416 417 418 419 420 421

  # construct list of Hap/PAR boundaries for entire seq region
  my @bounds;
  my $ent_slice = $sa->fetch_by_seq_region_id($sr_id);
  $ent_slice    = $ent_slice->invert() if($slice->strand == -1);
  my @ent_proj  = @{$sa->fetch_normalized_slice_projection($ent_slice)};

  shift @ent_proj; # skip first
  @bounds = map {$_->from_start - $slice->start() + 1} @ent_proj;


422
  # fetch features for the primary slice AND all symlinked slices
423 424 425
  foreach my $seg (@proj) {
    my $offset = $seg->from_start();
    my $seg_slice  = $seg->to_Slice();
426

427
    my $features = $self->_slice_fetch($seg_slice, $constraint);
428

429 430
    # if this was a symlinked slice offset the feature coordinates as needed
    if($seg_slice->name() ne $slice->name()) {
431 432

    FEATURE:
433 434
      foreach my $f (@$features) {
        if($offset != 1) {
435

436 437 438
          $f->{'start'} += $offset-1;
          $f->{'end'}   += $offset-1;
        }
439 440 441 442 443 444 445 446

        # discard boundary crossing features from symlinked regions
        foreach my $bound (@bounds) {
          if($f->{'start'} < $bound && $f->{'end'} >= $bound) {
            next FEATURE;
          }
        }

447 448 449 450 451 452 453 454 455
        $f->{'slice'} = $slice;
        push @result, $f;
      }
    } else {
      push @result, @$features;
    }
  }

  $self->{'_slice_feature_cache'}->{$key} = \@result;
456

457 458
  return \@result;
}
459 460


461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493
#
# helper function used by fetch_all_by_Slice_constraint method
#
sub _slice_fetch {
  my $self = shift;
  my $slice = shift;
  my $orig_constraint = shift;

  my $slice_start  = $slice->start();
  my $slice_end    = $slice->end();
  my $slice_strand = $slice->strand();
  my $slice_cs     = $slice->coord_system();
  my $slice_seq_region = $slice->seq_region_name();

  #get the synonym and name of the primary_table
  my @tabs = $self->_tables;
  my ($tab_name, $tab_syn) = @{$tabs[0]};

  #find out what coordinate systems the features are in
  my $mcc = $self->db->get_MetaCoordContainer();
  my @feat_css = @{$mcc->fetch_all_CoordSystems_by_feature_type($tab_name)};

  my $asma = $self->db->get_AssemblyMapperAdaptor();
  my @features;

  # fetch the features from each coordinate system they are stored in
 COORD_SYSTEM: foreach my $feat_cs (@feat_css) {
    my $mapper;
    my @coords;
    my @ids;

    if($feat_cs->equals($slice_cs)) {
      # no mapping is required if this is the same coord system
494

495
      my $max_len = $self->_max_feature_length() ||
496 497
        $mcc->fetch_max_length_by_CoordSystem_feature_type($feat_cs,$tab_name);

498 499 500 501 502 503
      my $constraint = $orig_constraint;

      my $sr_id = $self->db->get_SliceAdaptor->get_seq_region_id($slice);
      $constraint .= " AND " if($constraint);
      $constraint .=
          "${tab_syn}.seq_region_id = $sr_id AND " .
504 505
          "${tab_syn}.seq_region_start <= $slice_end AND " .
          "${tab_syn}.seq_region_end >= $slice_start";
506 507 508 509 510 511 512

      if($max_len) {
        my $min_start = $slice_start - $max_len;
        $constraint .=
          " AND ${tab_syn}.seq_region_start >= $min_start";
      }

513
      my $fs = $self->generic_fetch($constraint,undef,$slice);
514

515 516
      # features may still have to have coordinates made relative to slice
      # start
Graham McVicker's avatar
Graham McVicker committed
517
      $fs = _remap($fs, $mapper, $slice);
518

519 520 521
      push @features, @$fs;
    } else {
      $mapper = $asma->fetch_by_CoordSystems($slice_cs, $feat_cs);
522

523 524 525 526
      # Get list of coordinates and corresponding internal ids for
      # regions the slice spans
      @coords = $mapper->map($slice_seq_region, $slice_start, $slice_end,
                             $slice_strand, $slice_cs);
527

528
      @coords = grep {!$_->isa('Bio::EnsEMBL::Mapper::Gap')} @coords;
529

530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546
      next COORD_SYSTEM if(!@coords);

      @ids = map {$_->id()} @coords;
      @ids = @{$asma->seq_regions_to_ids($feat_cs, \@ids)};

      # When regions are large and only partially spanned by slice
      # it is faster to to limit the query with start and end constraints.
      # Take simple approach: use regional constraints if there are less
      # than a specific number of regions covered.

      if(@coords > $MAX_SPLIT_QUERY_SEQ_REGIONS) {
        my $constraint = $orig_constraint;
        my $id_str = join(',', @ids);
        $constraint .= " AND " if($constraint);
        $constraint .= "${tab_syn}.seq_region_id IN ($id_str)";

        my $fs = $self->generic_fetch($constraint, $mapper, $slice);
547

Graham McVicker's avatar
Graham McVicker committed
548
        $fs = _remap($fs, $mapper, $slice);
549 550 551 552 553

        push @features, @$fs;

      } else {
        # do multiple split queries using start / end constraints
554

555
        my $max_len = $self->_max_feature_length() ||
556 557
          $mcc->fetch_max_length_by_CoordSystem_feature_type($feat_cs,
                                                             $tab_name);
558 559 560 561 562
        my $len = @coords;
        for(my $i = 0; $i < $len; $i++) {
          my $constraint = $orig_constraint;
          $constraint .= " AND " if($constraint);
          $constraint .=
563 564 565
              "${tab_syn}.seq_region_id = "     . $ids[$i] . " AND " .
              "${tab_syn}.seq_region_start <= " . $coords[$i]->end() . " AND ".
              "${tab_syn}.seq_region_end >= "   . $coords[$i]->start();
566 567 568 569 570 571 572

          if($max_len) {
            my $min_start = $coords[$i]->start() - $max_len;
            $constraint .=
              " AND ${tab_syn}.seq_region_start >= $min_start";
          }

573
          my $fs = $self->generic_fetch($constraint,$mapper,$slice);
574

Graham McVicker's avatar
Graham McVicker committed
575
          $fs = _remap($fs, $mapper, $slice);
576

577
          push @features, @$fs;
578 579 580
        }
      }
    }
581
  } #COORD system loop
582

583
  return \@features;
584
}
585

586

587

588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605
#
# Helper function containing some common feature storing functionality
#
# Given a Feature this will return a copy (or the same feature if no changes 
# to the feature are needed) of the feature which is relative to the start
# of the seq_region it is on. The seq_region_id of the seq_region it is on
# is also returned.
#
# This method will also ensure that the database knows which coordinate
# systems that this feature is stored in.
#

sub _pre_store {
  my $self    = shift;
  my $feature = shift;

  if(!ref($feature) || !$feature->isa('Bio::EnsEMBL::Feature')) {
    throw('Expected Feature argument.');
606
  }
607

608 609 610
  $self->_check_start_end_strand($feature->start(),$feature->end(),
                                 $feature->strand());

611 612 613 614

  my $db = $self->db();

  my $slice_adaptor = $db->get_SliceAdaptor();
615 616 617 618 619 620
  my $slice = $feature->slice();

  if(!ref($slice) || !$slice->isa('Bio::EnsEMBL::Slice')) {
    throw('Feature must be attached to Slice to be stored.');
  }

621
  # make sure feature coords are relative to start of entire seq_region
622
  if($slice->start != 1 || $slice->strand != 1) {
623
    #move feature onto a slice of the entire seq_region
624 625 626 627 628 629 630 631 632 633 634 635 636 637 638
    $slice = $slice_adaptor->fetch_by_region($slice->coord_system->name(),
                                             $slice->seq_region_name(),
                                             undef, #start
                                             undef, #end
                                             undef, #strand
                                             $slice->coord_system->version());

    $feature = $feature->transfer($slice);

    if(!$feature) {
      throw('Could not transfer Feature to slice of ' .
            'entire seq_region prior to storing');
    }
  }

639
  # Ensure this type of feature is known to be stored in this coord system.
640 641 642 643 644
  my $cs = $slice->coord_system;

  my ($tab) = $self->_tables();
  my $tabname = $tab->[0];

645 646 647
  my $mcc = $db->get_MetaCoordContainer();

  $mcc->add_feature_type($cs, $tabname);
648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695

  my $seq_region_id = $slice_adaptor->get_seq_region_id($slice);

  if(!$seq_region_id) {
    throw('Feature is associated with seq_region which is not in this DB.');
  }

  return ($feature, $seq_region_id);
}


#
# helper function used to validate start/end/strand and 
# hstart/hend/hstrand etc.
#
sub _check_start_end_strand {
  my $self = shift;
  my $start = shift;
  my $end   = shift;
  my $strand = shift;

  #
  # Make sure that the start, end, strand are valid
  #
  if(int($start) != $start) {
    throw("Invalid Feature start [$start].  Must be integer.");
  }
  if(int($end) != $end) {
    throw("Invalid Feature end [$end]. Must be integer.");
  }
  if(int($strand) != $strand || $strand < -1 || $strand > 1) {
    throw("Invalid Feature strand [$strand]. Must be -1, 0 or 1.");
  }
  if($end < $start) {
    throw("Invalid Feature start/end [$start/$end]. Start must be less " .
          "than or equal to end.");
  }

  return 1;
}


#
# Given a list of features checks if they are in the correct coord system
# by looking at the first features slice.  If they are not then they are
# converted and placed on the slice.
#
sub _remap {
Graham McVicker's avatar
Graham McVicker committed
696
  my ($features, $mapper, $slice) = @_;
697 698 699 700 701

  #check if any remapping is actually needed
  if(@$features && (!$features->[0]->isa('Bio::EnsEMBL::Feature') ||
                    $features->[0]->slice == $slice)) {
    return $features;
702
  }
703

704
  #remapping has not been done, we have to do our own conversion from
705 706 707 708 709 710 711 712 713 714
  #to slice coords

  my @out;

  my $slice_start = $slice->start();
  my $slice_end   = $slice->end();
  my $slice_strand = $slice->strand();
  my $slice_cs    = $slice->coord_system();

  my ($seq_region, $start, $end, $strand);
715

716
  my $slice_seq_region = $slice->seq_region_name();
717

718
  foreach my $f (@$features) {
719
    #since feats were obtained in contig coords, attached seq is a contig
720 721 722 723 724 725
    my $fslice = $f->slice();
    if(!$fslice) {
      throw("Feature does not have attached slice.\n");
    }
    my $fseq_region = $fslice->seq_region_name();
    my $fcs = $fslice->coord_system();
726

727 728
    if(!$slice_cs->equals($fcs)) {
      #slice of feature in different coord system, mapping required
729

730 731 732 733 734 735 736 737 738 739 740 741 742
      ($seq_region, $start, $end, $strand) =
        $mapper->fastmap($fseq_region,$f->start(),$f->end(),$f->strand(),$fcs);

      # undefined start means gap
      next if(!defined $start);
    } else {
      $start      = $f->start();
      $end        = $f->end();
      $strand     = $f->strand();
      $seq_region = $f->slice->seq_region_name();
    }

    # maps to region outside desired area
743 744
    next if ($start > $slice_end) || ($end < $slice_start) || 
      ($slice_seq_region ne $seq_region);
745

746
    #shift the feature start, end and strand in one call
747
    if($slice_strand == -1) {
748
      $f->move( $slice_end - $end + 1, $slice_end - $start + 1, $strand * -1 );
749 750 751
    } else {
      $f->move( $start - $slice_start + 1, $end - $slice_start + 1, $strand );
    }
752 753 754

    $f->slice($slice);

Web Admin's avatar
Web Admin committed
755
    push @out,$f;
756
  }
757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806

  return \@out;
}


#
# Given a logic name and an existing constraint this will
# add an analysis table constraint to the feature.  Note that if no
# analysis_id exists in the columns of the primary table then no
# constraint is added at all
#
sub _logic_name_to_constraint {
  my $self = shift;
  my $constraint = shift;
  my $logic_name = shift;

  return $constraint if(!$logic_name);

  #make sure that an analysis_id exists in the primary table
  my ($prim_tab) = $self->_tables();
  my $prim_synonym = $prim_tab->[1];

  my $found_analysis=0;
  foreach my $col ($self->_columns) {
    my ($syn,$col_name) = split(/\./,$col);
    next if($syn ne $prim_synonym);
    if($col_name eq 'analysis_id') {
      $found_analysis = 1;
      last;
    }
  }

  if(!$found_analysis) {
    warning("This feature is not associated with an analysis.\n" .
            "Ignoring logic_name argument = [$logic_name].\n");
    return $constraint;
  }

  my $aa = $self->db->get_AnalysisAdaptor();
  my $an = $aa->fetch_by_logic_name($logic_name);

  if(!$an) {
    return undef;
  }

  my $an_id = $an->dbID();

  $constraint .= ' AND' if($constraint);
  $constraint .= " ${prim_synonym}.analysis_id = $an_id";
  return $constraint;
807 808 809
}


810
=head2 store
811

812
  Arg [1]    : list of Bio::EnsEMBL::SeqFeature
813 814 815 816 817 818 819
  Example    : $adaptor->store(@feats);
  Description: ABSTRACT  Subclasses are responsible for implementing this 
               method.  It should take a list of features and store them in 
               the database.
  Returntype : none
  Exceptions : thrown method is not implemented by subclass
  Caller     : general
820 821 822 823 824 825

=cut

sub store{
  my $self = @_;

826
  throw("Abstract method store not defined by implementing subclass\n");
827 828 829
}


830 831 832 833 834 835 836 837 838
=head2 remove

  Arg [1]    : A feature $feature 
  Example    : $feature_adaptor->remove($feature);
  Description: This removes a feature from the database.  The table the
               feature is removed from is defined by the abstract method
               _tablename, and the primary key of the table is assumed
               to be _tablename() . '_id'.  The feature argument must 
               be an object implementing the dbID method, and for the
839
               feature to be removed from the database a dbID value must
840 841
               be returned.
  Returntype : none
842 843
  Exceptions : thrown if $feature arg does not implement dbID(), or if
               $feature->dbID is not a true value
844 845 846 847
  Caller     : general

=cut

848

849 850 851
sub remove {
  my ($self, $feature) = @_;

852 853
  if(!$feature || !ref($feature) || !$feature->isa('Bio::EnsEMBL::Feature')) {
    throw('Feature argument is required');
854 855
  }

856 857
  if(!$feature->is_stored($self->db)) {
    throw("This feature is not stored in this database");
858 859
  }

860 861
  my @tabs = $self->_tables;
  my ($table) = @{$tabs[0]};
862 863 864 865

  my $sth = $self->prepare("DELETE FROM $table WHERE ${table}_id = ?");
  $sth->execute($feature->dbID());

866 867 868 869
  #unset the feature dbID ad adaptor
  $feature->dbID(undef);
  $feature->adaptor(undef);

870 871 872 873
  return;
}


874
=head2 remove_by_Slice
875

876 877 878 879 880 881 882 883
  Arg [1]    : Bio::Ensembl::Slice $slice
  Example    : $feature_adaptor->remove_by_RawContig($slice);
  Description: This removes features from the database which lie on a region
               represented by the passed in slice.  Only features which are
               fully contained by the slice are deleted; features which overlap
               the edge of the slice are not removed.
               The table the features are removed from is defined by
               the abstract method_tablename.
884
  Returntype : none
885
  Exceptions : thrown if no slice is supplied
886 887 888 889
  Caller     : general

=cut

890 891
sub remove_by_Slice {
  my ($self, $slice) = @_;
892

893 894
  if(!$slice || !ref($slice) || !$slice->isa('Bio::EnsEMBL::Slice')) {
    throw("Slice argument is required");
895 896
  }

897 898
  my @tabs = $self->_tables;
  my ($table_name) = @{$tabs[0]};
899

900 901 902
  my $seq_region_id = $self->db->get_SliceAdaptor->get_seq_region_id($slice);
  my $start = $slice->start();
  my $end   = $slice->end();
903

904 905 906 907 908 909 910
  #
  # Delete only features fully on the slice, not overlapping ones
  #
  my $sth = $self->prepare("DELETE FROM $table_name " .
                           "WHERE seq_region_id = ? " .
                           "AND   seq_region_start >= ? " .
                           "AND   seq_region_end <= ?");
911

912 913
  $sth->execute($seq_region_id, $start, $end);
  $sth->finish();
914 915 916 917
}



918 919


920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935
#_tables
#
#  Args       : none
#  Example    : $tablename = $self->_table_name()
#  Description: ABSTRACT PROTECTED Subclasses are responsible for implementing
#               this method.  It should list of [tablename, alias] pairs.  
#               Additionally the primary table (with the dbID, analysis_id, and
#               score) should be the first table in the list.
#               e.g:
#               ( ['repeat_feature',   'rf'],
#                 ['repeat_consensus', 'rc']);
#               used to obtain features.  
#  Returntype : list of [tablename, alias] pairs
#  Exceptions : thrown if not implemented by subclass
#  Caller     : BaseFeatureAdaptor::generic_fetch
#
936

937
sub _tables {
938 939
  my $self = shift;

940
  throw("abstract method _tables not defined by implementing" .
Simon Potter's avatar
Simon Potter committed
941
               " subclass of BaseFeatureAdaptor");
942 943 944
  return undef;
}

945

946 947 948 949 950 951 952 953 954 955 956
#_columns
#
#  Args       : none
#  Example    : $tablename = $self->_columns()
#  Description: ABSTRACT PROTECTED Subclasses are responsible for implementing
#               this method.  It should return a list of columns to be used
#               for feature creation
#  Returntype : list of strings
#  Exceptions : thrown if not implemented by subclass
#  Caller     : BaseFeatureAdaptor::generic_fetch
#
957 958 959 960

sub _columns {
  my $self = shift;

961
  throw("abstract method _columns not defined by implementing" .
Simon Potter's avatar
Simon Potter committed
962
               " subclass of BaseFeatureAdaptor");
963 964 965
}


966

967 968 969 970 971 972 973 974 975 976 977 978
# _default_where_clause
#
#  Arg [1]    : none
#  Example    : none
#  Description: May be overridden to provide an additional where constraint to 
#               the SQL query which is generated to fetch feature records.
#               This constraint is always appended to the end of the generated
#               where clause
#  Returntype : string
#  Exceptions : none
#  Caller     : generic_fetch
#
979

980 981
sub _default_where_clause {
  my $self = shift;
982