SliceAdaptor.pm 74.4 KB
Newer Older
1
=head1 LICENSE
2

Andy Yates's avatar
Andy Yates committed
3
  Copyright (c) 1999-2012 The European Bioinformatics Institute and
4
  Genome Research Limited.  All rights reserved.
5

6 7 8 9 10 11 12 13
  This software is distributed under a modified Apache license.
  For license details, please see

    http://www.ensembl.org/info/about/code_licence.html

=head1 CONTACT

  Please email comments or questions to the public Ensembl
14
  developers list at <dev@ensembl.org>.
15 16 17 18 19

  Questions may also be sent to the Ensembl help desk at
  <helpdesk@ensembl.org>.

=cut
20 21 22

=head1 NAME

23 24
Bio::EnsEMBL::DBSQL::SliceAdaptor - A database aware adaptor responsible for
the creation of Slice objects.
25 26 27

=head1 SYNOPSIS

Graham McVicker's avatar
Graham McVicker committed
28
  use Bio::EnsEMBL::Utils::Slice qw(split_Slices);
29 30 31
  use Bio::EnsEMBL::Registry;

  Bio::EnsEMBL::Registry->load_registry_from_db(
32 33
    -host => 'ensembldb.ensembl.org',
    -user => 'anonymous'
34 35
  );

36 37
  $slice_adaptor =
    Bio::EnsEMBL::Registry->get_adaptor( "human", "core", "slice" );
38

Graham McVicker's avatar
Graham McVicker committed
39
  # get a slice on the entire chromosome X
40
  $chr_slice = $slice_adaptor->fetch_by_region( 'chromosome', 'X' );
Graham McVicker's avatar
Graham McVicker committed
41 42

  # get a slice for each clone in the database
43 44
  foreach $cln_slice ( @{ $slice_adaptor->fetch_all('clone') } ) {
    # do something with clone
45 46
  }

Graham McVicker's avatar
Graham McVicker committed
47
  # get a slice which is part of NT_004321
48 49 50
  $spctg_slice =
    $slice_adaptor->fetch_by_region( 'supercontig', 'NT_004321',
    200_000, 600_000 );
51

52 53
  # get all non-redundant slices from the highest possible coordinate
  # systems
Graham McVicker's avatar
Graham McVicker committed
54 55 56
  $slices = $slice_adaptor->fetch_all('toplevel');

  # include non-reference regions
57
  $slices = $slice_adaptor->fetch_all( 'toplevel', undef, 1 );
Graham McVicker's avatar
Graham McVicker committed
58 59

  # include non-duplicate regions
60
  $slices = $slice_adaptor->fetch_all( 'toplevel', undef, 0, 1 );
Graham McVicker's avatar
Graham McVicker committed
61 62

  # split up a list of slices into smaller slices
63
  $overlap    = 1000;
Graham McVicker's avatar
Graham McVicker committed
64
  $max_length = 1e6;
65
  $slices     = split_Slices( $slices, $max_length, $overlap );
Graham McVicker's avatar
Graham McVicker committed
66 67

  # store a list of slice names in a file
68
  open( FILE, ">$filename" ) or die("Could not open file $filename");
Graham McVicker's avatar
Graham McVicker committed
69 70 71 72 73 74
  foreach my $slice (@$slices) {
    print FILE $slice->name(), "\n";
  }
  close FILE;

  # retreive a list of slices from a file
75 76
  open( FILE, $filename ) or die("Could not open file $filename");
  while ( $name = <FILE> ) {
Graham McVicker's avatar
Graham McVicker committed
77 78 79 80 81
    chomp($name);
    $slice = $slice_adaptor->fetch_by_name($name);
    # do something with slice
  }

82 83
=head1 DESCRIPTION

84 85
This module is responsible for fetching Slices representing genomic
regions from a database.  A Details on how slices can be used are in the
86
Bio::EnsEMBL::Slice module.
87

88
=head1 METHODS
89 90 91 92

=cut


93
package Bio::EnsEMBL::DBSQL::SliceAdaptor;
94 95 96
use vars qw(@ISA);
use strict;

97

98 99
use Bio::EnsEMBL::DBSQL::BaseAdaptor;
use Bio::EnsEMBL::Slice;
100
use Bio::EnsEMBL::CircularSlice;
101
use Bio::EnsEMBL::Mapper;
Ian Longden's avatar
Ian Longden committed
102
use Bio::EnsEMBL::LRGSlice;
103
use Bio::EnsEMBL::Utils::Exception qw(throw deprecate warning stack_trace_dump);
Andy Yates's avatar
Andy Yates committed
104
use Scalar::Util qw/looks_like_number/;
105

106
@ISA = ('Bio::EnsEMBL::DBSQL::BaseAdaptor');
107

108 109
sub new {
  my $caller = shift;
110

111
  my $class = ref($caller) || $caller;
112

113
  my $self = $class->SUPER::new(@_);
114

115 116
  # use a cache which is shared and also used by the assembly
  # mapper adaptor
117

118 119 120 121
  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'};
122

Ian Longden's avatar
Ian Longden committed
123 124 125
  $self->{'lrg_region_test'} = undef;
  my $meta_container = $self->db->get_MetaContainer();
  my @values = $meta_container->list_value_by_key("LRG");
126
  if(scalar(@values) and $values[0]->[0]){
Ian Longden's avatar
Ian Longden committed
127 128
    $self->{'lrg_region_test'} = $values[0]->[0];
  }
129 130 131 132 133 134
  return $self;
}


=head2 fetch_by_region

135
  Arg [1]    : string $coord_system_name (optional)
136
               The name of the coordinate system of the slice to be created
137
               This may be a name of an actual coordinate system or an alias
138 139 140 141
               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
142
               created on.
143 144 145 146 147 148 149 150
  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)
151 152
  Arg [7]    : boolean $no_fuzz (optional, default = undef (false))
               If true (non-zero), do not use "fuzzy matching" (see below).
153 154 155
  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
156 157 158 159 160 161 162 163 164 165 166
               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.
167 168 169 170 171 172 173

               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'.

174 175 176
               The fuzzy matching can be turned off by setting the
               $no_fuzz argument to a true value.

177 178 179 180 181 182 183
               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
184
  Caller     : general
185
  Status     : Stable
186 187 188

=cut

189 190 191 192

#
# ARNE: This subroutine needs simplification!! 
#
193
sub fetch_by_region {
194 195 196
  my ( $self, $coord_system_name, $seq_region_name, $start, $end,
       $strand, $version, $no_fuzz )
    = @_;
197

198 199
  if ( !defined($start) )  { $start  = 1 }
  if ( !defined($strand) ) { $strand = 1 }
200

201 202 203
  if ( !defined($seq_region_name) ) {
    throw('seq_region_name argument is required');
  }
204

205 206
  my $cs;
  my $csa = $self->db->get_CoordSystemAdaptor();
207

208 209
  if ( defined($coord_system_name) ) {
    $cs = $csa->fetch_by_name( $coord_system_name, $version );
210

211
    ## REMOVE THESE THREE LINES WHEN STICKLEBACK DB IS FIXED!
212 213 214 215 216 217 218
    ## Anne/ap5 (2007-10-09):
    # The problem was that the stickleback genebuild called the
    # chromosomes 'groups', which meant they weren't being picked out by
    # the karyotype drawing code.  Apparently they are usually called
    # 'groups' in the stickleback community, even though they really are
    # chromosomes!

219 220
    if ( !defined($cs) && $coord_system_name eq 'chromosome' ) {
      $cs = $csa->fetch_by_name( 'group', $version );
221 222
    }

223
    if ( !defined($cs) ) {
224
      throw( sprintf( "Unknown coordinate system:\n"
225
                        . "name='%s' version='%s'\n",
226
                      $coord_system_name, $version ) );
227
    }
228

229
    # fetching by toplevel is same as fetching w/o name or version
230 231
    if ( $cs->is_top_level() ) {
      $cs      = undef;
232 233
      $version = undef;
    }
234

235
  } ## end if ( defined($coord_system_name...))
236 237 238

  my $constraint;
  my $sql;
239
  my @bind_params;
240 241
  my $key;

242 243 244 245
  if ( defined($cs) ) {
    $sql = sprintf( "SELECT sr.name, sr.seq_region_id, sr.length, %d "
                      . "FROM seq_region sr ",
                    $cs->dbID() );
246

247 248
    $constraint = "AND sr.coord_system_id = ?";
    push( @bind_params, [ $cs->dbID(), SQL_INTEGER ] );
249

250
    $key = "$seq_region_name:" . $cs->dbID();
251
  } else {
252 253 254
    $sql =
      "SELECT sr.name, sr.seq_region_id, sr.length, cs.coord_system_id "
      . "FROM seq_region sr, coord_system cs ";
255

256 257 258 259 260
    $constraint = "AND sr.coord_system_id = cs.coord_system_id "
      . "AND cs.species_id = ? ";
    push( @bind_params, [ $self->species_id(), SQL_INTEGER ] );

    if ( defined($version) ) {
261
      $constraint .= "AND cs.version = ? ";
262
      push( @bind_params, [ $version, SQL_VARCHAR ] );
263
    }
264

265
    $constraint .= "ORDER BY cs.rank ASC";
266 267
  }

268
  # check the cache so we only go to the db if necessary
269
  my $length;
270
  my $arr;
271

272 273 274
  if ( defined($key) ) { $arr = $self->{'sr_name_cache'}->{$key} }

  if ( defined($arr) ) {
275
    $length = $arr->[3];
276
  } else {
277 278 279 280 281 282 283 284
    my $sth =
      $self->prepare( $sql . "WHERE sr.name = ? " . $constraint );

    unshift( @bind_params, [ $seq_region_name, SQL_VARCHAR ] );

    my $pos = 0;
    foreach my $param (@bind_params) {
      $sth->bind_param( ++$pos, $param->[0], $param->[1] );
285
    }
286 287

    $sth->execute();
288

289
    if ( $sth->rows() == 0 ) {
290
      $sth->finish();
Graham McVicker's avatar
Graham McVicker committed
291

292 293

      # try synonyms
294
      my $syn_sql_sth = $self->prepare("select s.name, s.coord_system_id from seq_region s, seq_region_synonym ss where s.seq_region_id = ss.seq_region_id and ss.synonym = ?");
295 296 297
      $syn_sql_sth->bind_param(1, $seq_region_name, SQL_VARCHAR);
      $syn_sql_sth->execute();
      my $new_name;
298 299
      my $new_coord_system;
      $syn_sql_sth->bind_columns( \$new_name, \$new_coord_system);
300
            
301
      if($syn_sql_sth->fetch){
302
        $syn_sql_sth->finish;
303 304 305 306 307 308 309 310
        if (not defined($cs)) {
            return $self->fetch_by_region($new_coord_system, $new_name, $start, $end, $strand, $version, $no_fuzz);
        } elsif ($cs->dbID != $new_coord_system) {
            warning("Searched for a known feature on coordinate system: ".$cs->dbID." but found it on: ".$new_coord_system.
            "\n No result returned, consider searching without coordinate system or use toplevel.");
            return;
        }
        
311 312 313 314
      }
      $syn_sql_sth->finish;


315 316
      if ($no_fuzz) { return undef }

317 318
      # Do fuzzy matching, assuming that we are just missing a version
      # on the end of the seq_region name.
319

320 321
      $sth =
        $self->prepare( $sql . " WHERE sr.name LIKE ? " . $constraint );
322

323 324 325 326 327 328
      $bind_params[0] =
        [ sprintf( '%s.%%', $seq_region_name ), SQL_VARCHAR ];

      $pos = 0;
      foreach my $param (@bind_params) {
        $sth->bind_param( ++$pos, $param->[0], $param->[1] );
329
      }
330

331
      $sth->execute();
332 333

      my $prefix_len = length($seq_region_name) + 1;
334 335
      my $high_ver   = undef;
      my $high_cs    = $cs;
336

337 338
      # Find the fuzzy-matched seq_region with the highest postfix
      # (which ought to be a version).
339

340 341
      my ( $tmp_name, $id, $tmp_length, $cs_id );
      $sth->bind_columns( \( $tmp_name, $id, $tmp_length, $cs_id ) );
342

Felix Kokocinski's avatar
Felix Kokocinski committed
343
      my $i = 0;
344

345 346 347
      while ( $sth->fetch ) {
        my $tmp_cs =
          ( defined($cs) ? $cs : $csa->fetch_by_dbID($cs_id) );
348

349
        # cache values for future reference
350
        my $arr = [ $id, $tmp_name, $cs_id, $tmp_length ];
351
        $self->{'sr_name_cache'}->{"$tmp_name:$cs_id"} = $arr;
352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370
        $self->{'sr_id_cache'}->{"$id"}                = $arr;

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

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

        # 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
371
        }
372 373

        $i++;
374
      } ## end while ( $sth->fetch )
375 376
      $sth->finish();

377
      # warn if fuzzy matching found more than one result
378 379 380 381 382 383 384
      if ( $i > 1 ) {
        warning(
          sprintf(
            "Fuzzy matching of seq_region_name "
              . "returned more than one result.\n"
              . "You might want to check whether the returned seq_region\n"
              . "(%s:%s) is the one you intended to fetch.\n",
385
            $high_cs->name(), $seq_region_name ) );
386 387
      }

388 389
      $cs = $high_cs;

390 391
      # return if we did not find any appropriate match:
      if ( !defined($high_ver) ) { return undef }
392

393 394
    } else {

395 396 397
      my ( $id, $cs_id );
      ( $seq_region_name, $id, $length, $cs_id ) =
        $sth->fetchrow_array();
398
      $sth->finish();
399

400
      # cache to speed up for future queries
401
      my $arr = [ $id, $seq_region_name, $cs_id, $length ];
402
      $self->{'sr_name_cache'}->{"$seq_region_name:$cs_id"} = $arr;
403 404
      $self->{'sr_id_cache'}->{"$id"}                       = $arr;
      $cs = $csa->fetch_by_dbID($cs_id);
405
    }
406
  } ## end else [ if ( defined($arr) ) ]
407

408
  if ( !defined($end) ) { $end = $length }
Graham McVicker's avatar
Graham McVicker committed
409

410 411
  #If this was given then check if we've got a circular seq region otherwise
  #let it fall through to the normal Slice method
412
  if ( $end + 1 < $start ) {
413
    my $cs_id = $cs->dbID();
414
    my $seq_region_id = $self->{'sr_name_cache'}->{"$seq_region_name:$cs_id"}->[0];
415 416 417 418 419 420 421 422 423 424 425 426 427
    if($self->is_circular($seq_region_id)) {
      my $new_sl =
        Bio::EnsEMBL::CircularSlice->new(
                                     -COORD_SYSTEM    => $cs,
                                     -SEQ_REGION_NAME => $seq_region_name,
                                     -SEQ_REGION_LENGTH => $length,
                                     -START             => $start,
                                     -END               => $end,
                                     -STRAND            => 1,
                                     -ADAPTOR           => $self );
  
      return $new_sl;
    }
428 429
  }

430 431 432
  if ( defined( $self->{'lrg_region_test'} )
       and substr( $cs->name, 0, 3 ) eq $self->{'lrg_region_test'} )
  {
Ian Longden's avatar
Ian Longden committed
433
    return
434 435 436 437 438 439 440 441
      Bio::EnsEMBL::LRGSlice->new( -COORD_SYSTEM    => $cs,
                                   -SEQ_REGION_NAME => $seq_region_name,
                                   -SEQ_REGION_LENGTH => $length,
                                   -START             => $start,
                                   -END               => $end,
                                   -STRAND            => $strand,
                                   -ADAPTOR           => $self );
  } else {
Ian Longden's avatar
Ian Longden committed
442
    return
443 444 445 446 447 448 449 450
      Bio::EnsEMBL::Slice->new_fast( {
                                  'coord_system'    => $cs,
                                  'seq_region_name' => $seq_region_name,
                                  'seq_region_length' => $length,
                                  'start'             => $start,
                                  'end'               => $end,
                                  'strand'            => $strand,
                                  'adaptor'           => $self } );
Ian Longden's avatar
Ian Longden committed
451
  }
452
} ## end sub fetch_by_region
453

454 455 456
=head2 fetch_by_toplevel_location

  Arg [1]     : string $location
457
                Ensembl formatted location. Can be a format like 
Andy Yates's avatar
Andy Yates committed
458 459 460 461 462 463
                C<name:start-end>, C<name:start..end>, C<name:start:end>, 
                C<name:start>, C<name>. We can also support strand 
                specification as a +/- or 1/-1. 
                
                Location names must be separated by a C<:>. All others can be
                separated by C<..>, C<:> or C<->.
464 465
  Arg[2]      : boolean $no_warnings
                Suppress warnings from this method
466
  Example     : my $slice = $sa->fetch_by_toplevel_location('X:1-10000')
Andy Yates's avatar
Andy Yates committed
467
                my $slice = $sa->fetch_by_toplevel_location('X:1-10000:-1')
468 469 470
  Description : Converts an Ensembl location/region into the sequence region
                name, start and end and passes them onto C<fetch_by_region()>. 
                The code assumes that the required slice is on the top level
471 472
                coordinate system. The code assumes that location formatting
                is not perfect and will perform basic cleanup before parsing.
473 474 475 476 477 478 479 480
  Returntype  : Bio::EnsEMBL::Slice
  Exceptions  : If $location is false otherwise see C<fetch_by_region()>
  Caller      : General
  Status      : Beta

=cut

sub fetch_by_toplevel_location {
481
  my ($self, $location, $no_warnings) = @_;
482

483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521
  my ($seq_region_name, $start, $end, $strand) = $self->parse_location_to_values($location, $no_warnings);

  if(! $seq_region_name) {
    return;
  }
    
  if(defined $start && defined $end && $start > $end) {
    throw "Cannot request a slice whose start is greater than its end. Start: $start. End: $end";
  }
  
  my $coord_system_name = 'toplevel';
  my $slice = $self->fetch_by_region($coord_system_name, $seq_region_name, $start, $end, $strand, undef, 0);
  return unless $slice;
  
  my $srl = $slice->seq_region_length();
  my $name = $slice->seq_region_name();
  if(defined $start && $start > $srl) {
    throw "Cannot request a slice whose start ($start) is greater than $srl for $name.";
  }
  if(defined $end && $end > $srl) {
    warning "Requested end ($end) is greater than $srl for $name. Resetting to $srl" if ! $no_warnings;
    $slice->{end} = $srl;
  }
  
  return $slice;
}

=head2 parse_location_to_values

  Arg [1]     : string $location
                Ensembl formatted location. Can be a format like 
                C<name:start-end>, C<name:start..end>, C<name:start:end>, 
                C<name:start>, C<name>. We can also support strand 
                specification as a +/- or 1/-1. 
                
                Location names must be separated by a C<:>. All others can be
                separated by C<..>, C<:> or C<->.
  Arg[2]      : boolean $no_warnings
                Suppress warnings from this method
522 523
  Arg[3]      : boolean $no_errors
                Supress errors being thrown from this method
524 525 526 527 528 529 530 531 532
  Example			: my ($name, $start, $end, $strand) = $sa->parse_location_to_values('X:1..100:1);
  Description	: Takes in an Ensembl location String and returns the parsed
                values
  Returntype 	: List. Contains name, start, end and strand 

=cut


sub parse_location_to_values {
533
  my ($self, $location, $no_warnings, $no_errors) = @_;
534 535 536
  
  throw 'You must specify a location' if ! $location;
  
537
  #cleanup any nomenclature like 1_000 or 1 000 or 1,000
538
  my $number_seps_regex = qr/\s+|,|_/;
539 540 541
  my $separator_regex = qr/(?:-|[.]{2}|\:)?/;
  my $number_regex = qr/[0-9,_ E]+/xms;
  my $strand_regex = qr/[+-1]|-1/xms;
542
  
543
  my $regex = qr/^((?:\w|\.|_|-)+) \s* :? \s* ($number_regex)? $separator_regex ($number_regex)? $separator_regex ($strand_regex)? $/xms;
544 545
  my ($seq_region_name, $start, $end, $strand);
  if(($seq_region_name, $start, $end, $strand) = $location =~ $regex) {
Andy Yates's avatar
Andy Yates committed
546 547 548 549 550 551 552
    
    if(defined $strand) {
      if(!looks_like_number($strand)) {
        $strand = ($strand eq '+') ? 1 : -1;
      }
    }
    
553
    if(defined $start) {
554
      $start =~ s/$number_seps_regex//g; 
555 556 557 558
      if($start < 1) {
        warning "Start was less than 1 (${start}) which is not allowed. Resetting to 1"  if ! $no_warnings;
        $start = 1;
      }
559
    }
560
    if(defined $end) {
561
      $end =~ s/$number_seps_regex//g;
562
      if($end < 1) {
563
        throw "Cannot request negative or 0 end indexes through this interface. Given $end but expected something greater than 0" unless $no_errors;
564
      }
565 566 567
    }
    
    if(defined $start && defined $end && $start > $end) {
568
      throw "Cannot request a slice whose start is greater than its end. Start: $start. End: $end" unless $no_errors;
569 570
    }
  }
571 572
  
  return ($seq_region_name, $start, $end, $strand);
573 574
}

575
=head2 fetch_by_region_unique
576

577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598
  Arg [1]    : string $coord_system_name (optional)
               The name of the coordinate system of the slice to be created
               This may be a name of an actual coordinate system or an alias
               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
               created on.
  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)
  Arg [7]    : boolean $no_fuzz (optional, default = undef (false))
               If true (non-zero), do not use "fuzzy matching" (see below).
  Example    : $slice = $slice_adaptor->fetch_by_region_unique('chromosome', 'HSCHR6_MHC_COX');
  Description: Retrieves a slice on the requested region but returns only the unique
               parts of the slice.  At a minimum the
               name the name of the seq_region to fetch must be provided.
599

600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633
               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.

               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'.

               The fuzzy matching can be turned off by setting the
               $no_fuzz argument to a true value.

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

  Returntype : listref Bio::EnsEMBL::Slice
  Exceptions : throw if no seq_region_name is provided
               throw if invalid coord_system_name is provided
               throw if start > end is provided
  Caller     : general
  Status     : Stable

=cut

sub fetch_by_region_unique {
  my $self = shift;

634
  my @out   = ();
635
  my $slice = $self->fetch_by_region(@_);
636

637

638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655
  if ( !exists( $self->{'asm_exc_cache'} ) ) {
    $self->_build_exception_cache();
  }

  if ( exists(
          $self->{'asm_exc_cache'}->{ $self->get_seq_region_id($slice) }
       ) )
  {
    # 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, $segment->[2] );
656 657
      }
    }
658 659
  } else {
    @out = ($slice);
660
  }
661

662
  return \@out;
663
} ## end sub fetch_by_region_unique
664

665
=head2 fetch_by_name
666

Graham McVicker's avatar
Graham McVicker committed
667
  Arg [1]    : string $name
668 669 670 671 672 673 674 675 676
  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
677

678 679 680 681
               Returns undef if no seq_region with the provided name exists in
               the database.

  Returntype : Bio::EnsEMBL::Slice or undef
682 683
  Exceptions : throw if incorrent arg provided
  Caller     : Pipeline
684
  Status     : Stable
685 686 687

=cut

688 689 690
sub fetch_by_name {
  my $self = shift;
  my $name = shift;
Graham McVicker's avatar
Graham McVicker committed
691

692 693
  if(!$name) {
    throw("name argument is required");
Graham McVicker's avatar
Graham McVicker committed
694
  }
695 696 697

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

698
  if(scalar(@array) < 3 || scalar(@array) > 6) {
699
    throw("Malformed slice name [$name].  Format is " .
Laura Clarke's avatar
 
Laura Clarke committed
700
        "coord_system:version:name:start:end:strand");
701 702
  }

703
  # Rearrange arguments to suit fetch_by_region
704

705 706 707 708 709 710 711 712 713
  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
714
}
Graham McVicker's avatar
Graham McVicker committed
715 716 717



718 719 720 721 722
=head2 fetch_by_seq_region_id

  Arg [1]    : string $seq_region_id
               The internal identifier of the seq_region to create this slice
               on
723 724 725
  Arg [2]    : optional start
  Arg [3]    : optional end
  Arg [4]    : optional strand
726 727 728
  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.
729 730
               Returns undef if no such slice exists.
  Returntype : Bio::EnsEMBL::Slice or undef
Graham McVicker's avatar
Graham McVicker committed
731 732
  Exceptions : none
  Caller     : general
733
  Status     : Stable
Graham McVicker's avatar
Graham McVicker committed
734 735 736

=cut

737
sub fetch_by_seq_region_id {
738
  my ( $self, $seq_region_id, $start, $end, $strand ) = @_;
739

740 741
  my $arr = $self->{'sr_id_cache'}->{$seq_region_id};
  my ( $name, $length, $cs, $cs_id );
742

743

744 745
  if ( $arr && defined( $arr->[2] ) ) {
    ( $name, $cs_id, $length ) = ( $arr->[1], $arr->[2], $arr->[3] );
746
    $cs = $self->db->get_CoordSystemAdaptor->fetch_by_dbID($cs_id);
747
  } else {
748 749
    my $sth =
      $self->prepare(   "SELECT sr.name, sr.coord_system_id, sr.length "
750
                      . "FROM seq_region sr "
751
                      . "WHERE sr.seq_region_id = ? " );
752

753
    $sth->bind_param( 1, $seq_region_id, SQL_INTEGER );
754
    $sth->execute();
755

756
    if ( $sth->rows() == 0 ) { return undef }
757

758
    ( $name, $cs_id, $length ) = $sth->fetchrow_array();
759 760 761 762 763
    $sth->finish();

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

    #cache results to speed up repeated queries
764
    my $arr = [ $seq_region_id, $name, $cs_id, $length ];
765 766 767

    $self->{'sr_name_cache'}->{"$name:$cs_id"} = $arr;
    $self->{'sr_id_cache'}->{"$seq_region_id"} = $arr;
768 769
  }

770
  return
771 772 773 774 775 776 777 778
    Bio::EnsEMBL::Slice->new_fast({ 
	                      'coord_system'     => $cs,
                              'seq_region_name'  => $name,
                              'seq_region_length'=> $length,
                              'start'            => $start || 1,
                              'end'              => $end || $length,
                              'strand'           => $strand || 1,
                              'adaptor'           => $self} );
779
} ## end sub fetch_by_seq_region_id
Arne Stabenau's avatar
Arne Stabenau committed
780

Graham McVicker's avatar
Graham McVicker committed
781

782 783

=head2 get_seq_region_id
Graham McVicker's avatar
Graham McVicker committed
784

Arne Stabenau's avatar
Arne Stabenau committed
785
  Arg [1]    : Bio::EnsEMBL::Slice $slice
786 787 788 789 790 791 792 793 794 795
               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
796
  Status     : Stable
Arne Stabenau's avatar
Arne Stabenau committed
797 798

=cut
Graham McVicker's avatar
Graham McVicker committed
799

800 801 802
sub get_seq_region_id {
  my $self = shift;
  my $slice = shift;
Graham McVicker's avatar
Graham McVicker committed
803

Ian Longden's avatar
Ian Longden committed
804
  if(!$slice || !ref($slice) || !($slice->isa('Bio::EnsEMBL::Slice') or $slice->isa('Bio::EnsEMBL::LRGSlice'))) {
805 806
    throw('Slice argument is required');
  }
807
  
808
  my $seq_region_name = $slice->seq_region_name();
809
  my $key = $seq_region_name.":".$slice->coord_system->dbID();
810
  my $arr = $self->{'sr_name_cache'}->{"$key"};
811

812 813
  if( $arr ) {
    return $arr->[0];
814 815
  }

816
  my $cs_id = $slice->coord_system->dbID();
817 818 819 820 821 822

  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
823 824 825
  $sth->bind_param(1,"$seq_region_name",SQL_VARCHAR);
  $sth->bind_param(2,$cs_id,SQL_INTEGER);
  $sth->execute();
826 827 828

  if($sth->rows() != 1) {
    throw("Non existant or ambigous seq_region:\n" .
829 830 831
          "  coord_system=[$cs_id],\n" .
          "  name=[$seq_region_name],\n");

832 833 834 835 836 837
  }

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

  #cache information for future requests
838
  $arr = [ $seq_region_id, $seq_region_name, $cs_id, $length ];
839 840 841 842

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

843
  return $seq_region_id;
Arne Stabenau's avatar
Arne Stabenau committed
844
}
845 846


Arne Stabenau's avatar
Arne Stabenau committed
847

848 849 850 851 852 853 854 855 856
=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
857
  Arg [3]    : bool $include_non_reference (optional)
858 859 860
               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
861 862
  Arg [4]    : int $include_duplicates (optional)
               If set duplicate regions will be returned.
863
               
864
               NOTE: if you do not use this option and you have a PAR
865 866 867
               (pseudo-autosomal region) at the beginning of your seq_region
               then your slice will not start at position 1, so coordinates
               retrieved from this slice might not be what you expected.
868

869 870 871 872
  Arg[5]     : bool $include_lrg (optional)  (default 0)
               If set lrg regions will be returned aswell.


873 874 875
  Example    : @chromos = @{$slice_adaptor->fetch_all('chromosome','NCBI33')};
               @contigs = @{$slice_adaptor->fetch_all('contig')};

Graham McVicker's avatar
Graham McVicker committed
876 877 878 879 880
               # 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)};
881 882 883 884 885 886

  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.
887 888
               If the coordinate system with the provided name and version
               does not exist an empty list is returned.