Class: Bio::BFRTools::BFRRegion

Inherits:
DB::Fasta::Region
  • Object
show all
Defined in:
lib/bio/BFRTools.rb

Constant Summary collapse

BASES = 
[:A, :C, :G, :T]

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(opts) ⇒ BFRRegion

Returns a new instance of BFRRegion.



111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
 
# File 'lib/bio/BFRTools.rb', line 111

def initialize(opts)
  opts = { :min_cov=>20, :max_snp_1kbp => 5 }.merge!(opts)
  reg = Bio::DB::Fasta::Region.parse_region(opts[:region])
  self.entry = reg.entry
  self.start = reg.start
  self.end   = reg.end
  @BFRs = nil
  opts[:region] = reg
  @container = opts[:container]

  parental_1_reg = @container.parental_1_sam.fetch_region(opts)
  parental_2_reg = @container.parental_2_sam.fetch_region(opts)
  bulk_1_reg = @container.bulk_1_sam.fetch_region(opts)
  bulk_2_reg = @container.bulk_2_sam.fetch_region(opts)
  


  @parental_1_sequence = parental_1_reg.consensus
  @parental_2_sequence = parental_2_reg.consensus
  
  @bulk_1_sequence = bulk_1_reg.consensus
  @bulk_2_sequence = bulk_2_reg.consensus

  @snp_count = Container.snps_between( @parental_1_sequence , @parental_2_sequence )

  @ratios_bulk_1 = bulk_1_reg.base_ratios
  @ratios_bulk_2 = bulk_2_reg.base_ratios

  @bases_bulk_1 = bulk_1_reg.bases
  @bases_bulk_2 = bulk_2_reg.bases

  @avg_cov_bulk_1 = bulk_1_reg.average_coverage
  @avg_cov_bulk_2 = bulk_2_reg.average_coverage

  @coverages_1 =  bulk_1_reg.coverages
  @coverages_2 =  bulk_2_reg.coverages

end

Instance Attribute Details

#avg_cov_bulk_1 ⇒ Object (readonly)

Returns the value of attribute avg_cov_bulk_1.



109
110
111
 
# File 'lib/bio/BFRTools.rb', line 109

def avg_cov_bulk_1
  @avg_cov_bulk_1
end

#avg_cov_bulk_2 ⇒ Object (readonly)

Returns the value of attribute avg_cov_bulk_2.



109
110
111
 
# File 'lib/bio/BFRTools.rb', line 109

def avg_cov_bulk_2
  @avg_cov_bulk_2
end

#bases_bulk_1 ⇒ Object (readonly)

Returns the value of attribute bases_bulk_1.



109
110
111
 
# File 'lib/bio/BFRTools.rb', line 109

def bases_bulk_1
  @bases_bulk_1
end

#bases_bulk_2 ⇒ Object (readonly)

Returns the value of attribute bases_bulk_2.



109
110
111
 
# File 'lib/bio/BFRTools.rb', line 109

def bases_bulk_2
  @bases_bulk_2
end

#bulk_1_sequence ⇒ Object (readonly)

Returns the value of attribute bulk_1_sequence.



108
109
110
 
# File 'lib/bio/BFRTools.rb', line 108

def bulk_1_sequence
  @bulk_1_sequence
end

#bulk_2_sequence ⇒ Object (readonly)

Returns the value of attribute bulk_2_sequence.



108
109
110
 
# File 'lib/bio/BFRTools.rb', line 108

def bulk_2_sequence
  @bulk_2_sequence
end

#coverages_1 ⇒ Object (readonly)

Returns the value of attribute coverages_1.



109
110
111
 
# File 'lib/bio/BFRTools.rb', line 109

def coverages_1
  @coverages_1
end

#coverages_2 ⇒ Object (readonly)

Returns the value of attribute coverages_2.



109
110
111
 
# File 'lib/bio/BFRTools.rb', line 109

def coverages_2
  @coverages_2
end

#parental_1_sequence ⇒ Object (readonly)

Returns the value of attribute parental_1_sequence.



108
109
110
 
# File 'lib/bio/BFRTools.rb', line 108

def parental_1_sequence
  @parental_1_sequence
end

#parental_2_sequence ⇒ Object (readonly)

Returns the value of attribute parental_2_sequence.



108
109
110
 
# File 'lib/bio/BFRTools.rb', line 108

def parental_2_sequence
  @parental_2_sequence
end

#ratios_bulk_1 ⇒ Object (readonly)

Returns the value of attribute ratios_bulk_1.



109
110
111
 
# File 'lib/bio/BFRTools.rb', line 109

def ratios_bulk_1
  @ratios_bulk_1
end

#ratios_bulk_2 ⇒ Object (readonly)

Returns the value of attribute ratios_bulk_2.



109
110
111
 
# File 'lib/bio/BFRTools.rb', line 109

def ratios_bulk_2
  @ratios_bulk_2
end

#snp_count ⇒ Object (readonly)

Returns the value of attribute snp_count.



108
109
110
 
# File 'lib/bio/BFRTools.rb', line 108

def snp_count
  @snp_count
end

Instance Method Details

#base_count_for_base(base, base_matrix) ⇒ Object 



349
350
351
352
353
354
355
 
# File 'lib/bio/BFRTools.rb', line 349

def base_count_for_base(base, base_matrix)
  bases = Array.new
  for i in (0..base_matrix.size-1)
    bases << base_matrix[i][base]
  end
  bases
end

#base_ratios_for_base(base, ratios_matrix) ⇒ Object 



341
342
343
344
345
346
347
 
# File 'lib/bio/BFRTools.rb', line 341

def base_ratios_for_base(base, ratios_matrix)
  ratios = Array.new
  for i in (0..ratios_matrix.size-1)
    ratios << ratios_matrix[i][base]
  end
  ratios
end

#bfrs ⇒ Object 



193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
 
# File 'lib/bio/BFRTools.rb', line 193

def bfrs
  return @BFRs if @BFRs
  @BFRs = Hash.new
  
  [:first, :second].each do | reference |
    @BFRs[reference] = Hash.new
    BASES.each do |base|
      @BFRs[reference][base] = Array.new
    end
  end
  

  for i in (0..self.size-1)
    ratios_1 = @ratios_bulk_1[i]
    ratios_2 = @ratios_bulk_2[i]
    BASES.each do |base|
      
      if ratios_1[base] == 0 and ratios_2[base] == 0
        bfr1 = 0
        bfr2  = 0
      elsif ratios_1[base] == 0
       bfr1  = 0
       bfr2 = Float::INFINITY
      elsif ratios_2[base] == 0
       bfr1 = Float::INFINITY
        bfr2 = 0
        #bfr = Float::INFINITY
      else
        bfr1  =  ratios_1[base] / ratios_2[base]
        bfr2 =  ratios_2[base] / ratios_1[base]
      end
      @BFRs[:first][base] << bfr1
      @BFRs[:second][base] << bfr2
    end
  end
  @BFRs
end

#get_bfr_line(position, base, reference) ⇒ Object 



231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
 
# File 'lib/bio/BFRTools.rb', line 231

def get_bfr_line(position, base, reference)
  if(reference == :first)
    informative = @container.parental_1_name
    ref_base = @parental_2_sequence[position]
  elsif(reference == :second )
    informative = @container.parental_2_name
    ref_base = @parental_1_sequence[position]
  else
    raise BFRToolsException.new ("The reference for the line should be :first or :second, but was " + reference.to_s )
  end
  
  relative_position = self.start +  position 
  
  bfr = bfrs[reference][base][position]
  cov_1 = @coverages_1[position]
  cov_2 = @coverages_2[position]
  ratios_1 = @ratios_bulk_1[position][base]
  ratios_2 = @ratios_bulk_2[position][base]
  base_1_count = @bases_bulk_1[position][base.to_sym]
  base_2_count = @bases_bulk_2[position][base.to_sym]
  #puts "bases_1 #{@bases_bulk_1[position].to_s}"
  #puts "bases_2 #{@bases_bulk_2[position].to_s}"
  line = String.new
  line << @container.parental_1_name  << "\t" << @container.parental_2_name << "\t" <<  @container.bulk_1_name << "\t" << @container.bulk_2_name << "\t" << self.entry << "\t"
  line << ref_base  << "\t" << relative_position.to_s 
  line << "\t" << base.to_s << "\t" 
  line << bfr.round(2).to_s << "\t"  
  line << cov_1.to_s << "\t" << cov_2.to_s  << "\t" 
  line << informative 
  line << "\t" << ratios_1.round(2).to_s << "\t" << ratios_2.round(2).to_s
  line << "\t" << base_1_count.to_s << "\t" << base_2_count.to_s
  line
end

#get_bfr_lines(opts = {}) ⇒ Object 



150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
 
# File 'lib/bio/BFRTools.rb', line 150

def get_bfr_lines(opts = {})
 
  opts = { :min_cov=>20, :max_snp_1kbp => 5 }.merge!(opts)
  p opts.inspect
  region = self
  line  = String.new
  info = Array.new
  
  for i in (0..region.size-1)
    
    if region.coverages_1[i] > opts[:min_cov] and region.coverages_2[i] > opts[:min_cov]
      BASES.each do |base|

        info.clear
        if  Bio::NucleicAcid.is_valid( region.parental_1_sequence[i],  base.to_s  ) and 
          not  Bio::NucleicAcid.is_valid( region.parental_2_sequence[i],  base.to_s  )
          info << :first
        end

        if   Bio::NucleicAcid.is_valid( region.parental_2_sequence[i],  base.to_s  ) and 
          not Bio::NucleicAcid.is_valid( region.parental_1_sequence[i],  base.to_s  )
          info << :second
        end


        for informative in info
          l = region.get_bfr_line(i, base, informative)
          puts l << "\n"
          line << l << "\n"

          #     output.print  line , "\n"
        end
      end
    end
  end
  line
end

#snp_1kbp ⇒ Object 



189
190
191
 
# File 'lib/bio/BFRTools.rb', line 189

def snp_1kbp
  @snp_count.to_f * 1000 / self.size.to_f
end

#to_csv ⇒ Object 



311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
 
# File 'lib/bio/BFRTools.rb', line 311

def to_csv
  out = String.new
  out << "Parental 1," << @container.parental_1_name << "\n"
  out << "Parental 2," << @container.parental_2_name << "\n"
  out << "Bulk 1, " <<  @container.bulk_1_name << "\n"
  out << "Bulk 2," << @container.bulk_2_name << "\n"
  out << "Positions," << (1..self.size).to_a.join(",") << "\n"
  out << "Parental 1 consensus," << @parental_1_sequence .split(//).join(",") << "\n"
  out << "Parental 2 consensus," << @parental_2_sequence .split(//).join(",") << "\n"
  out << "Bulk 1 consensus," << @bulk_1_sequence .split(//).join(",") << "\n"
  out << "Bulk 1 coverage," << @coverages_1.join(",") << "\n"
  #  puts BASES
  BASES.each do |base|
    out << "Bases Bulk 1"<< base.to_s << "," <<  base_count_for_base(base, @bases_bulk_1).join(",") << "\n"
    out << "Ratios Bulk 1 " << base.to_s << "," << base_ratios_for_base(base, @ratios_bulk_1).join(",")  << "\n"
  end
  out << "Bulk 2 consensus," << @bulk_2_sequence .split(//).join(",") << "\n"
  out << "Bulk 2 coverage," << @coverages_2.join(",") << "\n"

  BASES.each do |base|
    out << "Bases Bulk 2   "<< base.to_s << "," <<  base_count_for_base(base, @bases_bulk_2).join(",") << "\n"
    out << "Ratios Bulk 2 " << base.to_s << "," << base_ratios_for_base(base, @ratios_bulk_2).join(",")  << "\n"
  end
  BASES.each do |base|
    out << "BFRs" << base.to_s << ","  << bfrs[:first][base].join(",") << "\n"
  end
  #     << "\t" << @container.bulk_2_name << "\t" << self.entry << "\t"
  out
end

#to_json(opts) ⇒ Object 



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
 
# File 'lib/bio/BFRTools.rb', line 274

def to_json (opts)
  #      puts JSON.dump self
  #      JSON.dump self
  #{}"{\"firstName\": \"John\"}"
  out = String.new
  out << "{" 
  out << "\"Parental_1\" : \"" << @container.parental_1_name << "\"\n"
  out << "\"Parental 2\" : \"" << @container.parental_2_name << "\"\n"
  out << "\"Bulk 1\" :  \"" <<  @container.bulk_1_name << "\"\n"
  out << "\"Bulk 2\" : \"" << @container.bulk_2_name << "\"\n"
  out << "\"Positions\" : " << (1..self.size).to_a.to_json << "\n" #TODO: Make this for any subsection, so we can subquery in case we are working on something bigger
  out << "\"Parental_1_consensus\":" << @parental_1_sequence .split(//).to_json << "\n"
  out << "\"Parental_2_consensus\":" << @parental_2_sequence .split(//).to_json << "\n"
  out << "\"Bulk_1_consensus\":" << @bulk_1_sequence .split(//).to_json << "\n"
  out << "\"Bulk_1_coverage\":" << @coverages_1.to_json << "\n"
  #  puts BASES
  
  BASES.each do |base|
    out << "\"Bases_Bulk_1"  << base.to_s << "\":" <<  base_count_for_base(base, @bases_bulk_1).join(",") << "\n"
    out << "\"Ratios_Bulk_1" << base.to_s << "\":" << base_ratios_for_base(base, @ratios_bulk_1).join(",")  << "\n"
  end
  out << "\"Bulk_2_consensus\":" << @bulk_2_sequence .split(//).join(",") << "\n"
  out << "\"Bulk_2_coverage\":" << @coverages_2.join(",") << "\n"

  BASES.each do |base|
    out << "\"Bases_Bulk_2"<< base.to_s << "\":" <<  base_count_for_base(base, @bases_bulk_2).join(",") << "\n"
    out << "\"Ratios_Bulk_2" << base.to_s << "\":" << base_ratios_for_base(base, @ratios_bulk_2).join(",")  << "\n"
  end
  BASES.each do |base|
    out << "\"BFR" << base.to_s << "\":"  << bfrs[:first][base].join(",") << "\n"
  end
  #     << "\t" << @container.bulk_2_name << "\t" << self.entry << "\t"
  out << "}"
  out

end

#to_multi_fasta ⇒ Object 



265
266
267
268
269
270
271
272
 
# File 'lib/bio/BFRTools.rb', line 265

def to_multi_fasta
  fasta_string = String.new
  fasta_string << ">"<< self.to_s << ":" << @container.parental_1_name << "\n" << @parental_1_sequence << "\n"
  fasta_string << ">"<< self.to_s << ":" << @container.parental_2_name << "\n" << @parental_2_sequence << "\n"
  fasta_string << ">"<< self.to_s << ":" << @container.bulk_1_name << "\n" << @bulk_1_sequence << "\n"
  fasta_string << ">"<< self.to_s << ":" << @container.bulk_2_name << "\n" << @bulk_2_sequence << "\n"
  fasta_string 
end