Module: Bio::EMBLDB::Common

Included in:
Bio::EMBL, UniProtKB
Defined in:
lib/bio/db/embl/common.rb

Constant Summary collapse

DELIMITER = 
"\n//\n"
RS = 
DELIMITER
TAGSIZE = 
5

Instance Method Summary collapse

Instance Method Details

#ac ⇒ Object Also known as: accessions

returns a Array of accession numbers in the AC lines.

AC Line

"AC   A12345; B23456;"
AC [AC1;]+

Accession numbers format:

1       2     3          4          5          6
[O,P,Q] [0-9] [A-Z, 0-9] [A-Z, 0-9] [A-Z, 0-9] [0-9]



99
100
101
102
103
104
105
106
107
108
 
# File 'lib/bio/db/embl/common.rb', line 99

def ac
  unless @data['AC']
    tmp = Array.new
    field_fetch('AC').split(/ /).each do |e|
      tmp.push(e.sub(/;/,''))
    end
    @data['AC'] = tmp
  end
  @data['AC']
end

#accession ⇒ Object

returns the first accession number in the AC lines



113
114
115
 
# File 'lib/bio/db/embl/common.rb', line 113

def accession
  ac[0]
end

#de ⇒ Object Also known as: description, definition

returns a String int the DE line.

DE Line



121
122
123
124
125
126
 
# File 'lib/bio/db/embl/common.rb', line 121

def de
  unless @data['DE']
    @data['DE'] = fetch('DE')
  end
  @data['DE']
end

#dr ⇒ Object

returns contents in the DR line.

  • Bio::EMBLDB::Common#dr -> [ <Database cross-reference Hash>* ]

where <Database cross-reference Hash> is:

  • Bio::EMBLDB::Common#dr {|k,v| }

DR Line; defabases cross-reference (>=0) a cross_ref pre one line

"DR  database_identifier; primary_identifier; secondary_identifier."



329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
 
# File 'lib/bio/db/embl/common.rb', line 329

def dr
  unless @data['DR']
    tmp = Hash.new
    self.get('DR').split(/\n/).each do |db|
      a = db.sub(/^DR   /,'').sub(/.$/,'').strip.split(/;[ ]/)
      dbname = a.shift
      tmp[dbname] = Array.new unless tmp[dbname]
      tmp[dbname].push(a)
    end
    @data['DR'] = tmp
  end
  if block_given?
    @data['DR'].each do |k,v|
      yield(k, v)
    end
  else
    @data['DR']
  end
end

#initialize(entry) ⇒ Object 



86
87
88
 
# File 'lib/bio/db/embl/common.rb', line 86

def initialize(entry)
  super(entry, TAGSIZE)
end

#kw ⇒ Object Also known as: keywords

returns keywords in the KW line.

  • Bio::EMBLDB::Common#kw -> [ <keyword>* ]

KW Line; keyword (>=1)

KW   [Keyword;]+



220
221
222
223
224
225
226
227
228
229
230
 
# File 'lib/bio/db/embl/common.rb', line 220

def kw
  unless @data['KW']
    if get('KW').size > 0
      tmp = fetch('KW').sub(/.$/,'')
      @data['KW'] = tmp.split(/;/).map {|e| e.strip }
    else
      @data['KW'] = []
    end
  end
  @data['KW']
end

#oc ⇒ Object

returns contents in the OC line.

  • Bio::EMBLDB::Common#oc -> [ <organism class String>* ]

OC Line; organism classification (>=1)

OC   Eukaryota; Alveolata; Apicomplexa; Piroplasmida; Theileriidae;
OC   Theileria.



203
204
205
206
207
208
209
210
211
212
213
214
 
# File 'lib/bio/db/embl/common.rb', line 203

def oc
  unless @data['OC']
    begin
      @data['OC'] = fetch('OC').sub(/.$/,'').split(/;/).map {|e|
        e.strip 
      }
    rescue NameError
      nil
    end
  end
  @data['OC']
end

#og ⇒ Object

returns contents in the OG line.

  • Bio::EMBLDB::Common#og -> [ <ogranella String>* ]

OG Line; organella (0 or 1/entry)

OG   Plastid; Chloroplast.
OG   Mitochondrion.
OG   Plasmid sym pNGR234a.
OG   Plastid; Cyanelle.
OG   Plasmid pSymA (megaplasmid 1).
OG   Plasmid pNRC100, Plasmid pNRC200, and Plasmid pHH1.



180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
 
# File 'lib/bio/db/embl/common.rb', line 180

def og
  unless @data['OG']
    og = Array.new
    if get('OG').size > 0
      ogstr = fetch('OG')
      ogstr.sub!(/\.$/,'')
      ogstr.sub!(/ and/,'')
      ogstr.sub!(/;/, ',')
      ogstr.split(',').each do |tmp|
        og.push(tmp.strip)
      end
    end
    @data['OG'] = og
  end
  @data['OG']
end

#os(num = nil) ⇒ Object

returns contents in the OS line.

  • Bio::EMBLDB#os -> Array of <OS Hash>

where <OS Hash> is:

[{'name'=>'Human', 'os'=>'Homo sapiens'}, 
 {'name'=>'Rat', 'os'=>'Rattus norveticus'}]

  • Bio::SPTR#os[‘name’] => “Human”

  • Bio::SPTR#os => ‘os’=>‘Homo sapiens’

  • Bio::STPR#os(0) => “Homo sapiens (Human)”

OS Line; organism species (>=1)

"OS   Trifolium repens (white clover)"

OS   Genus species (name).
OS   Genus species (name0) (name1).
OS   Genus species (name0) (name1).
OS   Genus species (name0), G s0 (name0), and G s (name1).



148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
 
# File 'lib/bio/db/embl/common.rb', line 148

def os(num = nil)
  unless @data['OS']
    os = Array.new
    fetch('OS').split(/, and|, /).each do |tmp|
      if tmp =~ /([A-Z][a-z]* *[\w \:\'\+\-]+\w)/
        org = $1
        tmp =~ /(\(.+\))/ 
        os.push({'name' => $1, 'os' => org})
      else
        raise "Error: OS Line. #{$!}\n#{fetch('OS')}\n"
      end
    end
    @data['OS'] = os
  end
  if num
    # EX. "Trifolium repens (white clover)"
    "#{@data['OS'][num]['os']} {#data['OS'][num]['name']"
  end
  @data['OS']
end

#ref ⇒ Object

returns contents in the R lines.

  • Bio::EMBLDB::Common#ref -> [ <refernece information Hash>* ]

where <reference information Hash> is:

{'RN' => '', 'RC' => '', 'RP' => '', 'RX' => '', 
 'RA' => '', 'RT' => '', 'RL' => '', 'RG' => ''}

R Lines

  • RN RC RP RX RA RT RL RG



242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
 
# File 'lib/bio/db/embl/common.rb', line 242

def ref
  unless @data['R']
    ary = Array.new
    get('R').split(/\nRN   /).each do |str|
      raw = {'RN' => '', 'RC' => '', 'RP' => '', 'RX' => '', 
             'RA' => '', 'RT' => '', 'RL' => '', 'RG' => ''}
      str = 'RN   ' + str unless /^RN   / =~ str
      str.split("\n").each do |line|
        if /^(R[NPXARLCTG])   (.+)/ =~ line
          raw[$1] += $2 + ' '
        else
          raise "Invalid format in R lines, \n[#{line}]\n"
        end
      end
      raw.each_value {|v| 
        v.strip! 
        v.sub!(/^"/,'')
        v.sub!(/;$/,'')
        v.sub!(/"$/,'')
      }
      ary.push(raw)
    end
    @data['R'] = ary
  end
  @data['R']
end

#references ⇒ Object

returns Bio::Reference object from Bio::EMBLDB::Common#ref.

  • Bio::EMBLDB::Common#ref -> Bio::References



271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
 
# File 'lib/bio/db/embl/common.rb', line 271

def references
  unless @data['references']
    ary = self.ref.map {|ent|
      hash = Hash.new
      ent.each {|key, value|
        case key
        when 'RN'
          if /\[(\d+)\]/ =~ value.to_s
            hash['embl_gb_record_number'] = $1.to_i
          end
        when 'RC'
          unless value.to_s.strip.empty?
            hash['comments'] ||= []
            hash['comments'].push value
          end
        when 'RP'
          hash['sequence_position'] = value
        when 'RA'
          a = value.split(/\, /)
          a.each do |x|
            x.sub!(/( [^ ]+)\z/, ",\\1")
          end
          hash['authors'] = a
        when 'RT'
          hash['title'] = value
        when 'RL'
          if /(.*) (\d+) *(\(([^\)]+)\))?(\, |\:)([a-zA-Z\d]+\-[a-zA-Z\d]+) *\((\d+)\)\.?\z/ =~ value.to_s
            hash['journal'] = $1.rstrip
            hash['volume']  = $2
            hash['issue']   = $4
            hash['pages']   = $6
            hash['year']    = $7
          else
            hash['journal'] = value
          end
        when 'RX'  # PUBMED, DOI, (AGRICOLA)
          value.split(/\. /).each {|item|
            tag, xref = item.split(/\; /).map {|i| i.strip.sub(/\.\z/, '') }
            hash[ tag.downcase ]  = xref
          }
        end
      }
      Reference.new(hash)
    }
    @data['references'] = ary.extend(Bio::References::BackwardCompatibility)
  end
  @data['references']
end