Bio/DB/Universal.pm

   1
   2 #
   3 # BioPerl module for Bio::DB::Universal
   4 #
   5 # Please direct questions and support issues to <bioperl-l@bioperl.org>
   6 #
   7 # Cared for by Ewan Birney <birney@ebi.ac.uk>
   8 #
   9 # Copyright Ewan Birney
  10 #
  11 # You may distribute this module under the same terms as perl itself
  12
  13 # POD documentation - main docs before the code
  14
  15 =head1 NAME
  16
  17 Bio::DB::Universal - Artificial database that delegates to specific databases
  18
  19 =head1 SYNOPSIS
  20
  21     $uni = Bio::DB::Universal->new();
  22
  23     # by default connects to web databases. We can also
  24     # substitute local databases
  25
  26     $embl = Bio::Index::EMBL->new( -filename => '/some/index/filename/locally/stored');
  27     $uni->use_database('embl',$embl);
  28
  29     # treat it like a normal database. Recognises strings
  30     # like gb|XXXXXX and embl:YYYYYY
  31
  32     $seq1 = $uni->get_Seq_by_id("embl:HSHNRNPA");
  33     $seq2 = $uni->get_Seq_by_acc("gb|A000012");
  34
  35     # with no separator, tries to guess database. In this case the
  36     # _ is considered to be indicative of swissprot
  37     $seq3 = $uni->get_Seq_by_id('ROA1_HUMAN');
  38
  39 =head1 DESCRIPTION
  40
  41 Artificial database that delegates to specific databases, with a
  42 "smart" (well, smartish) guessing routine for what the ids. No doubt
  43 the smart routine can be made smarter.
  44
  45 The hope is that you can make this database and just throw ids at it -
  46 for most easy cases it will sort you out. Personally, I would be
  47 making sure I knew where each id came from and putting it into its own
  48 database first - but this is a quick and dirty solution.
  49
  50 By default this connects to web orientated databases, with all the
  51 reliability and network bandwidth costs this implies. However you can
  52 subsistute your own local databases - they could be Bio::Index
  53 databases (DBM file and flat file) or bioperl-db based (MySQL based)
  54 or biocorba-based (whatever you like behind the corba interface).
  55
  56 Internally the tags for the databases are
  57
  58    genbank - ncbi dna database
  59    embl    - ebi's dna database (these two share accession number space)
  60    swiss   - swissprot + sptrembl (EBI's protein database)
  61
  62 We should extend this for RefSeq and other sequence databases which
  63 are out there... ;)
  64
  65 Inspired by Lincoln Stein, written by Ewan Birney.
  66
  67 =head1 FEEDBACK
  68
  69 =head2 Mailing Lists
  70
  71 User feedback is an integral part of the evolution of this and other
  72 Bioperl modules. Send your comments and suggestions preferably to one
  73 of the Bioperl mailing lists.  Your participation is much appreciated.
  74
  75   bioperl-l@bio.perl.org
  76
  77 =head2 Support
  78
  79 Please direct usage questions or support issues to the mailing list:
  80
  81 I<bioperl-l@bioperl.org>
  82
  83 rather than to the module maintainer directly. Many experienced and
  84 reponsive experts will be able look at the problem and quickly
  85 address it. Please include a thorough description of the problem
  86 with code and data examples if at all possible.
  87
  88 =head2 Reporting Bugs
  89
  90 Report bugs to the Bioperl bug tracking system to help us keep track
  91 the bugs and their resolution.  Bug reports can be submitted via the
  92 web:
  93
  94   https://github.com/bioperl/bioperl-live/issues
  95
  96 =head1 AUTHOR - Ewan Birney
  97
  98 Email birney@ebi.ac.uk
  99
 100 =head1 APPENDIX
 101
 102 The rest of the documentation details each of the object
 103 methods. Internal methods are usually preceded with a _
 104
 105 =cut
 106
 107
 108 # Let the code begin...
 109
 110
 111 package Bio::DB::Universal;
 112 use strict;
 113
 114 # Object preamble - inherits from Bio::Root::Root
 115
 116
 117 use Bio::DB::GenBank;
 118 use Bio::DB::SwissProt;
 119 use Bio::DB::EMBL;
 120
 121
 122 use base qw(Bio::DB::RandomAccessI Bio::Root::Root);
 123 # new() can be inherited from Bio::Root::Root
 124
 125 sub new {
 126     my ($class) = @_;
 127
 128     my $self = {};
 129     bless $self,$class;
 130
 131     $self->{'db_hash'} = {};
 132
 133     # default databases
 134
 135     $self->use_database('embl',Bio::DB::EMBL->new);
 136     $self->use_database('genbank',Bio::DB::GenBank->new);
 137     $self->use_database('swiss',Bio::DB::GenBank->new);
 138
 139     return $self;
 140 }
 141
 142
 143 =head2 get_Seq_by_id
 144
 145  Title   : get_Seq_by_id
 146  Usage   :
 147  Function:
 148  Example :
 149  Returns :
 150  Args    :
 151
 152
 153 =cut
 154
 155 sub get_Seq_by_id{
 156    my ($self,$str) = @_;
 157
 158    my ($tag,$id) = $self->guess_id($str);
 159
 160    return $self->{'db_hash'}->{$tag}->get_Seq_by_id($id);
 161 }
 162
 163
 164 =head2 get_Seq_by_acc
 165
 166  Title   : get_Seq_by_acc
 167  Usage   :
 168  Function:
 169  Example :
 170  Returns :
 171  Args    :
 172
 173
 174 =cut
 175
 176 sub get_Seq_by_acc {
 177    my ($self,$str) = @_;
 178
 179    my ($tag,$id) = $self->guess_id($str);
 180
 181    return $self->{'db_hash'}->{$tag}->get_Seq_by_acc($id);
 182 }
 183
 184
 185
 186 =head2 guess_id
 187
 188  Title   : guess_id
 189  Usage   :
 190  Function:
 191  Example :
 192  Returns :
 193  Args    :
 194
 195
 196 =cut
 197
 198 sub guess_id{
 199    my ($self,$str) = @_;
 200
 201    if( $str =~ /(\S+)[:|\/;](\w+)/ ) {
 202        my $tag;
 203        my $db = $1;
 204        my $id = $2;
 205        if( $db =~ /gb/i || $db =~ /genbank/i || $db =~ /ncbi/i ) {
 206            $tag = 'genbank';
 207        } elsif ( $db =~ /embl/i || $db =~ /emblbank/ || $db =~ /^em/i ) {
 208            $tag = 'embl';
 209        } elsif ( $db =~ /swiss/i || $db =~ /^sw/i || $db =~ /sptr/ ) {
 210            $tag = 'swiss';
 211        } else {
 212            # throw for the moment
 213            $self->throw("Could not guess database type $db from $str");
 214        }
 215        return ($tag,$id);
 216
 217    } else {
 218        my $tag;
 219        # auto-guess from just the id
 220        if( $str =~ /_/ ) {
 221            $tag = 'swiss';
 222        } elsif ( $str =~ /^[QPR]\w+\d$/ ) {
 223            $tag = 'swiss';
 224        } elsif ( $str =~ /[A-Z]\d+/ ) {
 225            $tag = 'genbank';
 226        } else {
 227            # default genbank...
 228            $tag = 'genbank';
 229        }
 230        return ($tag,$str);
 231    }
 232
 233
 234 }
 235
 236
 237 =head2 use_database
 238
 239  Title   : use_database
 240  Usage   :
 241  Function:
 242  Example :
 243  Returns :
 244  Args    :
 245
 246
 247 =cut
 248
 249 sub use_database{
 250    my ($self,$name,$database) = @_;
 251
 252    $self->{'db_hash'}->{$name} = $database;
 253 }
 254
 255 1;