SliceAdaptor.pm 49.5 KB
Newer Older
1 2

#
Simon Potter's avatar
pod  
Simon Potter committed
3
# Ensembl module for Bio::EnsEMBL::DBSQL::SliceAdaptor
4 5
#
#
Graham McVicker's avatar
Graham McVicker committed
6
# Copyright (c) 2004 Ensembl
7 8 9 10 11 12 13
#
# You may distribute this module under the same terms as perl itself

# POD documentation - main docs before the code

=head1 NAME

14 15
Bio::EnsEMBL::DBSQL::SliceAdaptor - A database aware adaptor responsible for
the creation of Slice objects.
16 17 18

=head1 SYNOPSIS

Graham McVicker's avatar
Graham McVicker committed
19 20
  use Bio::EnsEMBL::DBSQL::DBAdaptor;
  use Bio::EnsEMBL::Utils::Slice qw(split_Slices);
21

Graham McVicker's avatar
Graham McVicker committed
22
  $db = Bio::EnsEMBL::DBSQL::DBAdaptor->new(...);
23

Graham McVicker's avatar
Graham McVicker committed
24
  $slice_adaptor = $db->get_SliceAdaptor();
25

Graham McVicker's avatar
Graham McVicker committed
26 27 28 29 30
  # get a slice on the entire chromosome X
  $chr_slice = $slice_adaptor->fetch_by_region('chromosome','X');

  # get a slice for each clone in the database
  foreach $cln_slice (@{$slice_adaptor->fetch_all('clone')}) {
31 32 33
    #do something with clone
  }

Graham McVicker's avatar
Graham McVicker committed
34 35
  # get a slice which is part of NT_004321
  $spctg_slice = $slice_adaptor->fetch_by_region('supercontig','NT_004321',
36
                                                    200_000, 600_000);
37 38


Graham McVicker's avatar
Graham McVicker committed
39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
  # get all non-redundant slices from the highest possible coordinate systems
  $slices = $slice_adaptor->fetch_all('toplevel');

  # include non-reference regions
  $slices = $slice_adaptor->fetch_all('toplevel',undef,1);

  # include non-duplicate regions
  $slices = $slice_adaptor->fetch_all('toplevel', undef, 0, 1);

  # split up a list of slices into smaller slices
  $overlap = 1000;
  $max_length = 1e6;
  $slices = split_Slices($slices, $max_length, $overlap);


  # store a list of slice names in a file
  open(FILE, ">$filename") or die("Could not open file $filename");
  foreach my $slice (@$slices) {
    print FILE $slice->name(), "\n";
  }
  close FILE;

  # retreive a list of slices from a file
  open(FILE, $filename) or die("Could not open file $filename");
  while($name = <FILE>) {
    chomp($name);
    $slice = $slice_adaptor->fetch_by_name($name);

    # do something with slice
  }

70 71
=head1 DESCRIPTION

72
This module is responsible for fetching Slices representing genomic regions
Graham McVicker's avatar
Graham McVicker committed
73
from a database.  A Details on how slices can be used are in the
74
Bio::EnsEMBL::Slice module.
75

76
=head1 CONTACT
77

Graham McVicker's avatar
Graham McVicker committed
78 79 80
This module is part of the Ensembl project http://www.ensembl.org

For more information email <ensembl-dev@ebi.ac.uk>
81

82
=head1 METHODS
83 84 85 86

=cut


87
package Bio::EnsEMBL::DBSQL::SliceAdaptor;
88 89 90
use vars qw(@ISA);
use strict;

91

92 93
use Bio::EnsEMBL::DBSQL::BaseAdaptor;
use Bio::EnsEMBL::Slice;
94
use Bio::EnsEMBL::Mapper;
95

96
use Bio::EnsEMBL::Utils::Exception qw(throw deprecate warning);
97

98

99
@ISA = ('Bio::EnsEMBL::DBSQL::BaseAdaptor');
100

101 102
sub new {
  my $caller = shift;
103

104
  my $class = ref($caller) || $caller;
105

106
  my $self = $class->SUPER::new(@_);
107

108 109 110 111 112 113
  # use a cache which is shared and also used by the assembly
  # mapper adaptor
  my $seq_region_cache = $self->db->get_SeqRegionCache();

  $self->{'sr_name_cache'} = $seq_region_cache->{'name_cache'};
  $self->{'sr_id_cache'}   = $seq_region_cache->{'id_cache'};
114 115 116 117 118 119 120

  return $self;
}


=head2 fetch_by_region

121
  Arg [1]    : string $coord_system_name (optional)
122
               The name of the coordinate system of the slice to be created
123
               This may be a name of an actual coordinate system or an alias
124 125 126 127
               to a coordinate system.  Valid aliases are 'seqlevel' or
               'toplevel'.
  Arg [2]    : string $seq_region_name
               The name of the sequence region that the slice will be
Graham McVicker's avatar
Graham McVicker committed
128
               created on.
129 130 131 132 133 134 135 136
  Arg [3]    : int $start (optional, default = 1)
               The start of the slice on the sequence region
  Arg [4]    : int $end (optional, default = seq_region length)
               The end of the slice on the sequence region
  Arg [5]    : int $strand (optional, default = 1)
               The orientation of the slice on the sequence region
  Arg [6]    : string $version (optional, default = default version)
               The version of the coordinate system to use (e.g. NCBI33)
137 138 139
  Example    : $slice = $slice_adaptor->fetch_by_region('chromosome', 'X');
               $slice = $slice_adaptor->fetch_by_region('clone', 'AC008066.4');
  Description: Retrieves a slice on the requested region.  At a minimum the
140 141 142 143 144 145 146 147 148 149 150
               name the name of the seq_region to fetch must be provided.

               If no coordinate system name is provided than a slice on the
               highest ranked coordinate system with a matching
               seq_region_name will be returned.  If a version but no
               coordinate system name is provided, the same behaviour will
               apply, but only coordinate systems of the appropriate version
               are considered.  The same applies if the 'toplevel' coordinate
               system is specified, however in this case the version is
               ignored.  The coordinate system should always be specified if
               it is known, since this is unambiguous and faster.
151 152 153 154 155 156 157 158 159 160 161 162 163 164

               Some fuzzy matching is performed if no exact match for
               the provided name is found.  This allows clones to be
               fetched even when their version is not known.  For
               example fetch_by_region('clone', 'AC008066') will
               retrieve the sequence_region with name 'AC008066.4'.

               If the requested seq_region is not found in the database undef
               is returned.

  Returntype : Bio::EnsEMBL::Slice or undef
  Exceptions : throw if no seq_region_name is provided
               throw if invalid coord_system_name is provided
               throw if start > end is provided
165
  Caller     : general
166
  Status     : Stable
167 168 169

=cut

170 171 172 173

#
# ARNE: This subroutine needs simplification!! 
#
174 175 176 177
sub fetch_by_region {
  my ($self, $coord_system_name, $seq_region_name,
      $start, $end, $strand, $version) = @_;

178 179
  $start  = 1 if (!defined($start));
  $strand = 1 if (!defined($strand));
180

181
  throw('seq_region_name argument is required') if (!defined($seq_region_name));
182

183 184
  my $cs;
  my $csa = $self->db->get_CoordSystemAdaptor();
185

186
  if ($coord_system_name) {
187 188
    $cs = $csa->fetch_by_name($coord_system_name,$version);

189 190 191 192
    if( !$cs && $coord_system_name eq 'chromosome' ) {
      $cs = $csa->fetch_by_name('group',$version);
    }

193
    if (!$cs) {
194 195 196 197 198
      throw("Unknown coordinate system:\n name='$coord_system_name' " .
            "version='$version'\n");
    }

    # fetching by toplevel is same as fetching w/o name or version
199
    if ($cs->is_top_level()) {
200 201 202
      $cs = undef;
      $version = undef;
    }
203
  }
204 205 206 207 208 209

  my $constraint;
  my $sql;
  my @bind_vals;
  my $key;

210
  if ($cs) {
211
    push @bind_vals, $cs->dbID();
212
    $sql = "SELECT sr.name, sr.seq_region_id, sr.length, " .
213 214 215 216
           $cs->dbID() ." FROM seq_region sr ";

    $constraint = "sr.coord_system_id = ?";

217
    $key = "$seq_region_name:".$cs->dbID();
218 219 220 221 222 223 224 225 226 227
  } else {
    $sql = "SELECT sr.name, sr.seq_region_id, sr.length, " .
           "       cs.coord_system_id " .
           "FROM   seq_region sr, coord_system cs ";

    $constraint = "sr.coord_system_id = cs.coord_system_id ";
    if($version) {
      $constraint .= "AND cs.version = ? ";
      push @bind_vals, $version;
    }
228
    $constraint .= "ORDER BY cs.rank ASC";
229 230
  }

231
  # check the cache so we only go to the db if necessary
232
  my $length;
233
  my $arr;
234
  if ($key) {
235
    $arr = $self->{'sr_name_cache'}->{$key};
236
  }
237

238
  if ($arr) {
239
    $length = $arr->[3];
240
  } else {
241
    my $sth = $self->prepare($sql . " WHERE sr.name = ? AND " . $constraint);
242 243 244
    # Quotes around "$seq_region_name" are needed so that mysql does not
    # treat chromosomes like '6' as an int.  This was doing horrible
    # inexact matches like '6DR51', '6_UN', etc.
245
    $sth->bind_param(1, "$seq_region_name", SQL_VARCHAR);
246
    if ($cs){
247
	$sth->bind_param(2, $cs->dbID, SQL_INTEGER);
248 249
    }
    else{
250
	$sth->bind_param(2, $version, SQL_VARCHAR)  if ($version);	
251
    }
252 253

    $sth->execute();
254

255
    if ($sth->rows() == 0) {
256
      $sth->finish();
Graham McVicker's avatar
Graham McVicker committed
257

258 259
      # do fuzzy matching, assuming that we are just missing a version on
      # the end of the seq_region name
260
      
261
      $sth = $self->prepare($sql . " WHERE sr.name LIKE ? AND " . $constraint);
262

263
      $sth->bind_param(1, "$seq_region_name.%", SQL_VARCHAR);
264 265

      if ($cs){
266
	  $sth->bind_param(2, $cs->dbID, SQL_INTEGER);
267 268
      }
      else{
269
	  $sth->bind_param(2, $version, SQL_VARCHAR) if ($version);	
270
      }
271
      $sth->execute();
272 273

      my $prefix_len = length($seq_region_name) + 1;
274 275
      my $high_ver = undef;
      my $high_cs = $cs;
276 277 278 279

      # find the fuzzy-matched seq_region with the highest postfix (which ought
      # to be a version)

280 281 282
      my ($tmp_name, $id, $tmp_length, $cs_id);
      $sth->bind_columns(\$tmp_name, \$id, \$tmp_length, \$cs_id);

283 284 285
      my $i;

      while ($sth->fetch) {
286
        my $tmp_cs = ($cs) ? $cs : $csa->fetch_by_dbID($cs_id);
287

288
        # cache values for future reference
289
        my $arr = [ $id, $tmp_name, $cs_id, $tmp_length ];
290 291
        $self->{'sr_name_cache'}->{"$tmp_name:$cs_id"} = $arr;
        $self->{'sr_id_cache'}->{"$id"} = $arr;
Graham McVicker's avatar
Graham McVicker committed
292

293
        my $tmp_ver = substr($tmp_name, $prefix_len);
294

295 296
        # skip versions which are non-numeric and apparently not versions
        next if($tmp_ver !~ /^\d+$/);
297

298 299 300 301 302 303 304 305
        # take version with highest num, if two versions match take one with
        # highest ranked coord system (lowest num)
        if(!defined($high_ver) || $tmp_ver > $high_ver ||
           ($tmp_ver == $high_ver && $tmp_cs->rank < $high_cs->rank)) {
            $seq_region_name = $tmp_name;
            $length          = $tmp_length;
            $high_ver        = $tmp_ver;
            $high_cs         = $tmp_cs;
Graham McVicker's avatar
Graham McVicker committed
306
        }
307 308

        $i++;
Graham McVicker's avatar
Graham McVicker committed
309
      }
310 311
      $sth->finish();

312 313 314 315 316
      # warn if fuzzy matching found more than one result
      if ($i > 1) {
        warning("Fuzzy matching of seq_region_name returned more than one result.\nYou might want to check whether the returned seq_region\n(".$high_cs->name.":$seq_region_name) is the one you intended to fetch.\n");
      }

317 318
      $cs = $high_cs;

319
      #return if we did not find any appropriate match:
320 321
      return undef if (!defined($high_ver));

322 323
    } else {

324 325
      my ($id, $cs_id);
      ($seq_region_name, $id, $length, $cs_id) = $sth->fetchrow_array();
326
      $sth->finish();
327

328 329
      # cahce to speed up for future queries
      my $arr = [ $id, $seq_region_name, $cs_id, $length ];
330 331
      $self->{'sr_name_cache'}->{"$seq_region_name:$cs_id"} = $arr;
      $self->{'sr_id_cache'}->{"$id"} = $arr;
332 333

      $cs = $csa->fetch_by_dbID( $cs_id );
334
    }
335
  }
336

337
  $end = $length if(!defined($end));
Graham McVicker's avatar
Graham McVicker committed
338

339 340
  if($end+1 < $start) {
    throw("start [$start] must be less than or equal to end+1 [$end+1]");
341 342
  }

343
  return Bio::EnsEMBL::Slice->new(-COORD_SYSTEM      => $cs,
344 345 346 347 348 349
                                  -SEQ_REGION_NAME   => $seq_region_name,
                                  -SEQ_REGION_LENGTH => $length,
                                  -START             => $start,
                                  -END               => $end,
                                  -STRAND            => $strand,
                                  -ADAPTOR           => $self);
350 351 352 353
}



354

355
=head2 fetch_by_name
356

Graham McVicker's avatar
Graham McVicker committed
357
  Arg [1]    : string $name
358 359 360 361 362 363 364 365 366
  Example    : $name  = 'chromosome:NCBI34:X:1000000:2000000:1';
               $slice = $slice_adaptor->fetch_by_name($name);
               $slice2 = $slice_adaptor->fetch_by_name($slice3->name());
  Description: Fetches a slice using a slice name (i.e. the value returned by
               the Slice::name method).  This is useful if you wish to 
               store a unique identifier for a slice in a file or database or
               pass a slice over a network.
               Slice::name allows you to serialise/marshall a slice and this
               method allows you to deserialise/unmarshal it.
Graham McVicker's avatar
Graham McVicker committed
367

368 369 370 371
               Returns undef if no seq_region with the provided name exists in
               the database.

  Returntype : Bio::EnsEMBL::Slice or undef
372 373
  Exceptions : throw if incorrent arg provided
  Caller     : Pipeline
374
  Status     : Stable
375 376 377

=cut

378 379 380
sub fetch_by_name {
  my $self = shift;
  my $name = shift;
Graham McVicker's avatar
Graham McVicker committed
381

382 383
  if(!$name) {
    throw("name argument is required");
Graham McVicker's avatar
Graham McVicker committed
384
  }
385 386 387

  my @array = split(/:/,$name);

388
  if(scalar(@array) < 3 || scalar(@array) > 6) {
389
    throw("Malformed slice name [$name].  Format is " .
Laura Clarke's avatar
 
Laura Clarke committed
390
        "coord_system:version:name:start:end:strand");
391 392
  }

393
  # Rearrange arguments to suit fetch_by_region
394

395 396 397 398 399 400 401 402 403
  my @targetarray;

  $targetarray[0]=$array[0];
  $targetarray[5]=(($array[1]&&$array[1] ne "")?$array[1]:undef);
  $targetarray[1]=(($array[2]&&$array[2] ne "")?$array[2]:undef);
  $targetarray[2]=(($array[3]&&$array[3] ne "")?$array[3]:undef);
  $targetarray[3]=(($array[4]&&$array[4] ne "")?$array[4]:undef);
  $targetarray[4]=(($array[5]&&$array[5] ne "")?$array[5]:undef);
  return $self->fetch_by_region(@targetarray);
Graham McVicker's avatar
Graham McVicker committed
404
}
Graham McVicker's avatar
Graham McVicker committed
405 406 407



408 409 410 411 412 413 414 415
=head2 fetch_by_seq_region_id

  Arg [1]    : string $seq_region_id
               The internal identifier of the seq_region to create this slice
               on
  Example    : $slice = $slice_adaptor->fetch_by_seq_region_id(34413);
  Description: Creates a slice object of an entire seq_region using the
               seq_region internal identifier to resolve the seq_region.
416 417
               Returns undef if no such slice exists.
  Returntype : Bio::EnsEMBL::Slice or undef
Graham McVicker's avatar
Graham McVicker committed
418 419
  Exceptions : none
  Caller     : general
420
  Status     : Stable
Graham McVicker's avatar
Graham McVicker committed
421 422 423

=cut

424 425 426
sub fetch_by_seq_region_id {
  my ($self, $seq_region_id) = @_;

427
  my $arr = $self->{'sr_id_cache'}->{ $seq_region_id };
428 429
  my ($name, $length, $cs);

430 431 432 433
  if( $arr ) {
    my $cs_id;
    ($name, $cs_id, $length ) = ( $arr->[1], $arr->[2], $arr->[3] );
    $cs = $self->db->get_CoordSystemAdaptor->fetch_by_dbID($cs_id);
434 435 436 437 438
  } else {
    my $sth = $self->prepare("SELECT name, length, coord_system_id " .
                             "FROM seq_region " .
                             "WHERE seq_region_id = ?");

439 440
    $sth->bind_param(1,$seq_region_id,SQL_INTEGER);
    $sth->execute();
441

442
    return undef if($sth->rows() == 0);
443 444 445 446 447 448 449 450

    my $cs_id;
    ($name, $length, $cs_id) = $sth->fetchrow_array();
    $sth->finish();

    $cs = $self->db->get_CoordSystemAdaptor->fetch_by_dbID($cs_id);

    #cache results to speed up repeated queries
451
    my $arr = [ $seq_region_id, $name, $cs_id, $length ];
452 453 454

    $self->{'sr_name_cache'}->{"$name:$cs_id"} = $arr;
    $self->{'sr_id_cache'}->{"$seq_region_id"} = $arr;
455 456
  }

457 458 459 460 461 462 463
  return Bio::EnsEMBL::Slice->new(-COORD_SYSTEM      => $cs,
                                  -SEQ_REGION_NAME   => $name,
                                  -SEQ_REGION_LENGTH => $length,
                                  -START             => 1,
                                  -END               => $length,
                                  -STRAND            => 1,
                                  -ADAPTOR           => $self);
Arne Stabenau's avatar
Arne Stabenau committed
464 465
}

Graham McVicker's avatar
Graham McVicker committed
466

467 468

=head2 get_seq_region_id
Graham McVicker's avatar
Graham McVicker committed
469

Arne Stabenau's avatar
Arne Stabenau committed
470
  Arg [1]    : Bio::EnsEMBL::Slice $slice
471 472 473 474 475 476 477 478 479 480
               The slice to fetch a seq_region_id for
  Example    : $srid = $slice_adaptor->get_seq_region_id($slice);
  Description: Retrieves the seq_region id (in this database) given a slice
               Seq region ids are not stored on the slices themselves
               because they are intended to be somewhat database independant
               and seq_region_ids vary accross databases.
  Returntype : int
  Exceptions : throw if the seq_region of the slice is not in the db
               throw if incorrect arg provided
  Caller     : BaseFeatureAdaptor
481
  Status     : Stable
Arne Stabenau's avatar
Arne Stabenau committed
482 483

=cut
Graham McVicker's avatar
Graham McVicker committed
484

485 486 487
sub get_seq_region_id {
  my $self = shift;
  my $slice = shift;
Graham McVicker's avatar
Graham McVicker committed
488

489 490 491
  if(!$slice || !ref($slice) || !$slice->isa('Bio::EnsEMBL::Slice')) {
    throw('Slice argument is required');
  }
492
  
493
  my $seq_region_name = $slice->seq_region_name();
494
  my $key = $seq_region_name.":".$slice->coord_system->dbID();
495
  my $arr = $self->{'sr_name_cache'}->{"$key"};
496

497 498
  if( $arr ) {
    return $arr->[0];
499 500
  }

501
  my $cs_id = $slice->coord_system->dbID();
502 503 504 505 506 507

  my $sth = $self->prepare("SELECT seq_region_id, length " .
                           "FROM seq_region " .
                           "WHERE name = ? AND coord_system_id = ?");

  #force seq_region_name cast to string so mysql cannot treat as int
508 509 510
  $sth->bind_param(1,"$seq_region_name",SQL_VARCHAR);
  $sth->bind_param(2,$cs_id,SQL_INTEGER);
  $sth->execute();
511 512 513

  if($sth->rows() != 1) {
    throw("Non existant or ambigous seq_region:\n" .
514 515 516
          "  coord_system=[$cs_id],\n" .
          "  name=[$seq_region_name],\n");

517 518 519 520 521 522
  }

  my($seq_region_id, $length) = $sth->fetchrow_array();
  $sth->finish();

  #cache information for future requests
523
  $arr = [ $seq_region_id, $seq_region_name, $cs_id, $length ];
524 525 526 527

  $self->{'sr_name_cache'}->{"$seq_region_name:$cs_id"} = $arr;
  $self->{'sr_id_cache'}->{"$seq_region_id"} = $arr;

528
  return $seq_region_id;
Arne Stabenau's avatar
Arne Stabenau committed
529
}
530 531


Arne Stabenau's avatar
Arne Stabenau committed
532

533 534 535 536 537 538 539 540 541
=head2 fetch_all

  Arg [1]    : string $coord_system_name
               The name of the coordinate system to retrieve slices of.
               This may be a name of an acutal coordinate system or an alias
               to a coordinate system.  Valid aliases are 'seqlevel' or
               'toplevel'.
  Arg [2]    : string $coord_system_version (optional)
               The version of the coordinate system to retrieve slices of
542
  Arg [3]    : bool $include_non_reference (optional)
543 544 545
               If this argument is not provided then only reference slices
               will be returned. If set, both reference and non refeference
               slices will be rerurned.
Graham McVicker's avatar
Graham McVicker committed
546 547
  Arg [4]    : int $include_duplicates (optional)
               If set duplicate regions will be returned.
548

549 550 551
  Example    : @chromos = @{$slice_adaptor->fetch_all('chromosome','NCBI33')};
               @contigs = @{$slice_adaptor->fetch_all('contig')};

Graham McVicker's avatar
Graham McVicker committed
552 553 554 555 556
               # get even non-reference regions
               @slices = @{$slice_adaptor->fetch_all('toplevel',undef,1)};

               # include duplicate regions (such as pseudo autosomal regions)
               @slices = @{$slice_adaptor->fetch_all('toplevel', undef,0,1)};
557 558 559 560 561 562

  Description: Retrieves slices of all seq_regions for a given coordinate
               system.  This is analagous to the methods fetch_all which were
               formerly on the ChromosomeAdaptor, RawContigAdaptor and
               CloneAdaptor classes.  Slices fetched span the entire
               seq_regions and are on the forward strand.
563 564
               If the coordinate system with the provided name and version
               does not exist an empty list is returned.
565
               If the coordinate system name provided is 'toplevel', all
566
               non-redundant toplevel slices are returned.
567 568 569 570

               Retrieved slices can be broken into smaller slices using the
               Bio::EnsEMBL::Utils::Slice module.

571
  Returntype : listref of Bio::EnsEMBL::Slices
572
  Exceptions : none
573
  Caller     : general
574
  Status     : Stable
Web Admin's avatar
Web Admin committed
575 576 577

=cut

578 579 580 581 582
sub fetch_all {
  my $self = shift;
  my $cs_name = shift;
  my $cs_version = shift || '';

583
  my ($include_non_reference, $include_duplicates) = @_;
Web Admin's avatar
Web Admin committed
584

585 586 587
  #
  # verify existance of requested coord system and get its id
  #
588 589 590 591 592 593
  my $csa       = $self->db->get_CoordSystemAdaptor();
  my $orig_cs   = $csa->fetch_by_name($cs_name, $cs_version);

  return [] if(!$orig_cs);

  my $sth;
594

595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613
  my %bad_vals=();
  #
  # Get a hash of non reference seq regions
  #
  if(!$include_non_reference){
    my $sth2 = $self->prepare(
			      "SELECT sr.seq_region_id ".
			      "FROM seq_region sr, seq_region_attrib sra, attrib_type at ".
			      " WHERE at.code='non_ref'".
			      "  AND sra.seq_region_id=sr.seq_region_id ".
			      "  AND at.attrib_type_id=sra.attrib_type_id " );
    $sth2->execute();
    my ($seq_region_id);
    $sth2->bind_columns(\$seq_region_id);
    while($sth2->fetch()) {
      $bad_vals{$seq_region_id} = 1;
    }
    $sth2->finish();
  }
614
  #
615
  # Retrieve the seq_regions from the database
616
  #
617
  if($orig_cs->is_top_level()) {
618 619 620 621 622 623 624 625
   $sth =
       $self->prepare("SELECT sr.seq_region_id, sr.name, sr.length, " .
                      "       sr.coord_system_id " .
                      "FROM seq_region sr, " .
                      "     seq_region_attrib sra, attrib_type at " .
                      "WHERE at.code='toplevel' " .
                      "AND   at.attrib_type_id=sra.attrib_type_id " .
                      "AND   sra.seq_region_id=sr.seq_region_id");
626 627
    $sth->execute();
  } else {
628 629 630 631
     $sth =
       $self->prepare('SELECT seq_region_id, name, length, coord_system_id ' .
                      'FROM   seq_region ' .
                      'WHERE  coord_system_id =?');
632 633
     $sth->bind_param(1,$orig_cs->dbID,SQL_INTEGER);
     $sth->execute();
634
  }
635

636 637
  my ($seq_region_id, $name, $length, $cs_id);
  $sth->bind_columns(\$seq_region_id, \$name, \$length, \$cs_id);
638

639
  my $cache_count = 0;
640 641 642

  my @out;
  while($sth->fetch()) {
643 644
    if(!defined($bad_vals{$seq_region_id})){
      my $cs = $csa->fetch_by_dbID($cs_id);
645

646
      if(!$cs) {
647
        throw("seq_region $name references non-existent coord_system $cs_id.");
648
      }
649

650 651
      #cache values for future reference, but stop adding to the cache once we
      #we know we have filled it up
652
      if($cache_count < $Bio::EnsEMBL::Utils::SeqRegionCache::SEQ_REGION_CACHE_SIZE) {
653
        my $arr = [ $seq_region_id, $name, $cs_id, $length ];
654 655 656

        $self->{'sr_name_cache'}->{"$name:$cs_id"} = $arr;
        $self->{'sr_id_cache'}->{"$seq_region_id"} = $arr;
657

658
        $cache_count++;
659
      }
660 661 662 663 664 665 666 667 668 669 670

      my $slice = Bio::EnsEMBL::Slice->new
        (-START             => 1,
         -END               => $length,
         -STRAND            => 1,
         -SEQ_REGION_NAME   => $name,
         -SEQ_REGION_LENGTH => $length,
         -COORD_SYSTEM      => $cs,
         -ADAPTOR           => $self);

      if(!defined($include_duplicates) or !$include_duplicates){
671 672 673 674 675 676 677 678 679 680 681 682
        # test if this slice *could* have a duplicate (exception) region
        $self->_build_exception_cache() if(!exists $self->{'asm_exc_cache'});
        if(exists $self->{asm_exc_cache}->{$seq_region_id}) {

          # Dereference symlinked assembly regions.  Take out
          # any regions which are symlinked because these are duplicates
          my @projection = @{$self->fetch_normalized_slice_projection($slice)};
          foreach my $segment ( @projection) {
            if($segment->[2]->seq_region_name() eq $slice->seq_region_name() &&
               $segment->[2]->coord_system->equals($slice->coord_system)) {
              push @out, $segment->[2];
            }
683
          }
684 685 686
        } else {
          # no duplicate regions
          push @out, $slice;
687
        }
688 689
      } else {
        # we want duplicates anyway so do not do any checks
690
        push @out, $slice;
691
      }
692
    }
693
  }
694
  $sth->finish();
695

696
  return \@out;
Web Admin's avatar
Web Admin committed
697 698 699
}


700

Web Admin's avatar
Web Admin committed
701 702 703 704 705 706
=head2 fetch_by_band

 Title   : fetch_by_band
 Usage   :
 Function: create a Slice representing a series of bands
 Example :
707
 Returns : Bio::EnsEMBL::Slice
Web Admin's avatar
Web Admin committed
708
 Args    : the band name
709
 Status     : Stable
Web Admin's avatar
Web Admin committed
710 711 712 713 714 715 716 717 718 719 720

=cut

sub fetch_by_band {
  my ($self,$band) = @_;

  my $sth = $self->db->prepare
        ("select s.name,max(k.seq_region_id)-min(k.seq_region_id, min(k.seq_region_start), max(k.seq_region_id) " .
         "from karyotype as k " .
         "where k.band like ? and k.seq_region_id = s.seq_region_id");

721 722
  $sth->bind_param(1,"$band%",SQL_VARCHAR);
  $sth->execute();
Web Admin's avatar
Web Admin committed
723 724 725 726 727 728 729 730 731 732
  my ( $seq_region_name, $discrepancy, $seq_region_start, $seq_region_end) = $sth->fetchrow_array;

  if($seq_region_name && $discrepancy>0) {
    throw("Band maps to multiple seq_regions");
  } else {
    return $self->fetch_by_region('toplevel',$seq_region_name,$seq_region_start,$seq_region_end);
  }
  throw("Band not recognised in database");
}

733 734 735 736 737 738
=head2 fetch_by_chr_band

 Title   : fetch_by_chr_band
 Usage   :
 Function: create a Slice representing a series of bands
 Example :
739
 Returns : Bio::EnsEMBL::Slice
740
 Args    : the band name
741
 Status     : Stable
742 743 744 745 746 747

=cut

sub fetch_by_chr_band {
  my ($self,$chr,$band) = @_;

Web Admin's avatar
Web Admin committed
748
  my $chr_slice = $self->fetch_by_region('toplevel', $chr);
749 750 751
  my $seq_region_id = $self->get_seq_region_id($chr_slice);

  my $sth = $self->db->prepare
Web Admin's avatar
Web Admin committed
752
        ("select min(k.seq_region_start), max(k.seq_region_end) " .
753 754 755
         "from karyotype as k " .
         "where k.seq_region_id = ? and k.band like ?");

756 757 758
  $sth->bind_param(1,$seq_region_id,SQL_INTEGER);
  $sth->bind_param(2,"$band%",SQL_VARCHAR);
  $sth->execute();
759 760 761
  my ( $slice_start, $slice_end) = $sth->fetchrow_array;

  if(defined $slice_start) {
Web Admin's avatar
Web Admin committed
762
    return $self->fetch_by_region('toplevel',$chr,$slice_start,$slice_end);
763 764 765
  }

  throw("Band not recognised in database");
766 767 768 769
}



Web Admin's avatar
Web Admin committed
770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785
=head2 fetch_by_exon_stable_id

  Arg [1]    : string $exonid
               The stable id of the exon around which the slice is 
               desired
  Arg [2]    : (optional) int $size
               The length of the flanking regions the slice should encompass 
               on either side of the exon (0 by default)
  Example    : $slc = $sa->fetch_by_exon_stable_id('ENSE00000302930',10);
  Description: Creates a slice around the region of the specified exon. 
               If a context size is given, the slice is extended by that 
               number of basepairs on either side of the
               transcript.
  Returntype : Bio::EnsEMBL::Slice
  Exceptions : Thrown if the exon is not in the database.
  Caller     : general
786
  Status     : Stable
Web Admin's avatar
Web Admin committed
787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802

=cut

sub fetch_by_exon_stable_id{
  my ($self,$exonid,$size) = @_;

  throw('Exon argument is required.') if(!$exonid);

  my $ea = $self->db->get_ExonAdaptor;
  my $exon = $ea->fetch_by_stable_id($exonid);

  throw("Exon [$exonid] does not exist in DB.") if(!$exon);

  return $self->fetch_by_Feature($exon, $size);
}

Graham McVicker's avatar
Graham McVicker committed
803
=head2 fetch_by_transcript_stable_id
804

Graham McVicker's avatar
Graham McVicker committed
805 806 807 808 809 810 811 812 813
  Arg [1]    : string $transcriptid
               The stable id of the transcript around which the slice is 
               desired
  Arg [2]    : (optional) int $size
               The length of the flanking regions the slice should encompass 
               on either side of the transcript (0 by default)
  Example    : $slc = $sa->fetch_by_transcript_stable_id('ENST00000302930',10);
  Description: Creates a slice around the region of the specified transcript. 
               If a context size is given, the slice is extended by that 
814 815
               number of basepairs on either side of the
               transcript.
Graham McVicker's avatar
Graham McVicker committed
816
  Returntype : Bio::EnsEMBL::Slice
817
  Exceptions : Thrown if the transcript is not in the database.
Graham McVicker's avatar
Graham McVicker committed
818
  Caller     : general
819
  Status     : Stable
820 821 822

=cut

Graham McVicker's avatar
Graham McVicker committed
823
sub fetch_by_transcript_stable_id{
824 825
  my ($self,$transcriptid,$size) = @_;

826 827
  throw('Transcript argument is required.') if(!$transcriptid);

828
  my $ta = $self->db->get_TranscriptAdaptor;
829
  my $transcript = $ta->fetch_by_stable_id($transcriptid);
830

831 832 833 834
  throw("Transcript [$transcriptid] does not exist in DB.") if(!$transcript);

  return $self->fetch_by_Feature($transcript, $size);
}
835

Graham McVicker's avatar
Graham McVicker committed
836 837


Graham McVicker's avatar
Graham McVicker committed
838 839
=head2 fetch_by_transcript_id

Graham McVicker's avatar
Graham McVicker committed
840 841 842 843 844 845 846
  Arg [1]    : int $transcriptid
               The unique database identifier of the transcript around which 
               the slice is desired
  Arg [2]    : (optional) int $size
               The length of the flanking regions the slice should encompass 
               on either side of the transcript (0 by default)
  Example    : $slc = $sa->fetch_by_transcript_id(24, 1000);
847 848 849 850
  Description: Creates a slice around the region of the specified transcript.
               If a context size is given, the slice is extended by that
               number of basepairs on either side of the
               transcript.
Graham McVicker's avatar
Graham McVicker committed
851
  Returntype : Bio::EnsEMBL::Slice
852 853
  Exceptions : throw on incorrect args
               throw if transcript is not in database
Graham McVicker's avatar
Graham McVicker committed
854
  Caller     : general
855
  Status     : Stable
856 857 858

=cut

Graham McVicker's avatar
Graham McVicker committed
859
sub fetch_by_transcript_id {
860
  my ($self,$transcriptid,$size) = @_;
Graham McVicker's avatar
Graham McVicker committed
861

862
  throw('Transcript id argument is required.') if(!$transcriptid);
Graham McVicker's avatar
Graham McVicker committed
863

864 865 866 867 868 869
  my $transcript_adaptor = $self->db()->get_TranscriptAdaptor();
  my $transcript = $transcript_adaptor->fetch_by_dbID($transcriptid);

  throw("Transcript [$transcriptid] does not exist in DB.") if(!$transcript);

  return $self->fetch_by_Feature($transcript, $size);
870 871
}

872 873


Graham McVicker's avatar
Graham McVicker committed
874
=head2 fetch_by_gene_stable_id
875

Graham McVicker's avatar
Graham McVicker committed
876
  Arg [1]    : string $geneid
877
               The stable id of the gene around which the slice is
Graham McVicker's avatar
Graham McVicker committed
878 879
               desired
  Arg [2]    : (optional) int $size
Graham McVicker's avatar
Graham McVicker committed
880
               The length of the flanking regions the slice should encompass
Graham McVicker's avatar
Graham McVicker committed
881 882
               on either side of the gene (0 by default)
  Example    : $slc = $sa->fetch_by_transcript_stable_id('ENSG00000012123',10);
Graham McVicker's avatar
Graham McVicker committed
883 884 885
  Description: Creates a slice around the region of the specified gene.
               If a context size is given, the slice is extended by that
               number of basepairs on either side of the gene.
886
               The slice will be created in the genes native coordinate system.
Graham McVicker's avatar
Graham McVicker committed
887
  Returntype : Bio::EnsEMBL::Slice
888 889
  Exceptions : throw on incorrect args
               throw if transcript does not exist
Graham McVicker's avatar
Graham McVicker committed
890
  Caller     : general
891
  Status     : Stable
892 893 894

=cut

895 896
sub fetch_by_gene_stable_id {
  my ($self,$geneid,$size) = @_;
897