Class: PluginExtractInserts

Inherits:
Plugin
  • Object
show all
Defined in:
lib/seqtrimnext/plugins/plugin_extract_inserts.rb

Overview

Author: Almudena Bocinos Rioboo

Defines the main methods that are necessary to execute PluginLinker

Inherit: Plugin

Instance Attribute Summary

Attributes inherited from Plugin

#stats

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from Plugin

#add_plugin_stats, #add_stats, #add_text_stats, auto_setup, #can_execute?, check_param, #do_blasts, #execute, get_graph_filename, get_graph_title, graph_ignored?, ignored_graphs, #initialize, #merge_hits, #overlapX?, valid_graphs

Constructor Details

This class inherits a constructor from Plugin

Class Method Details

.check_params(params) ⇒ Object

Returns an array with the errors due to parameters are missing



394
395
396
397
398
399
400
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 394

def self.check_params(params)
  errors=[]
  
  # self.check_param(errors,params,'min_insert_size_trimmed','Integer')
  
  return errors
end

.plot_setup(stats_value, stats_name, x, y, init_stats, plot) ⇒ Object 



402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 402

def self.plot_setup(stats_value,stats_name,x,y,init_stats,plot)

  # puts "============== #{stats_name}"
  
  # puts stats_name
  case stats_name
    
  when 'insert_size'
    plot.x_label= "Length"
    plot.y_label= "Count"
    plot.x_range="[0:#{init_stats['biggest_sequence_size'].to_i}]"
    # plot.x_range="[0:200]"
    
    plot.add_x(x)
    plot.add_y(y)
    
    plot.do_graph
    
    return true
  else
    return false
  end
  
end

Instance Method Details

#add_action_after_linker(overlap, actions, seq) ⇒ Object

It’s created an ActionInsert or ActionShortInsert after the ActionLinker 



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
149
150
151
152
153
154
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 122

def add_action_after_linker(overlap,actions,seq) 
                                                 
  # puts "INSERT2: [#{overlap.end_pos},#{seq.insert_end}]"
            
  insert_size = seq.insert_end-overlap.end_pos
  
  min_insert_size = @params.get_param('min_insert_size_trimmed').to_i

  # puts "INSERT_SIZE2 #{insert_size} > #{min_insert_size}"         
  
  # puts "INSERT2: [#{overlap.end_pos},#{seq.insert_start} #{seq.insert_end} #{seq.seq_fasta.size} #{seq.seq_fasta_orig.size}]"  

  if ((overlap.end_pos-seq.insert_start < seq.seq_fasta_orig.size-1) && (insert_size>=min_insert_size) )  #if overlap's positions are left 
    #It's created an ActionInsert after the Actionoverlap
    a = seq.new_action(overlap.end_pos-seq.insert_start+1,seq.seq_fasta.size-1,"ActionInsert") # adds the ActionInsert to the sequence  
    # puts " new after action #{overlap.end_pos} - #{seq.insert_start}" 
    # puts " 1---new insert despues del linker #{a.start_pos} #{a.end_pos} " 
    

    actions.push a
      
  elsif (overlap.end_pos-seq.insert_start<seq.seq_fasta_orig.size-1)   #if overlap's positions are right and insert's size is short 
    #It's created an ActionInsert after the ActionLinker
    a = seq.new_action(overlap.end_pos-seq.insert_start+1,seq.seq_fasta.size-1,"ActionShortInsert") # adds the ActionInsert to the sequence
    
    # puts " new after action #{overlap.end_pos} - #{seq.insert_start} +1" 
    # puts "2---new insert short despues del linker #{a.start_pos} #{a.end_pos} "
    actions.push a 
  end    
  # puts "#{a.start_pos} #{a.end_pos}" if !a.nil?
  @stats[:insert_size_right]={insert_size => 1}   
  
end

#add_action_before_linker(overlap, actions, seq) ⇒ Object

It’s created an ActionInsert or ActionShortInsert before the ActionLinker Used: in class PluginLinker and PluginMid



81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 81

def add_action_before_linker(overlap,actions,seq)
                                                
  # puts "INSERT1: [#{seq.insert_start},#{overlap.start_pos}]"
                                                        
  insert_size = overlap.start_pos - seq.insert_start
  
  min_insert_size = @params.get_param('min_insert_size_trimmed').to_i  
  
  # puts "INSERT1: [#{overlap.start_pos},#{seq.insert_start} #{seq.insert_end}   seqsize #{seq.seq_fasta.size} insert_size #{insert_size} #{min_insert_size}]" 
      
                                                 

  if ((overlap.start_pos>seq.insert_start) && (insert_size >= min_insert_size))#if overlap's positions are right    
      #It's created an ActionInsert or ActionShortInsert before the Actionoverlap    
      # a = seq.new_action(seq.insert_start_last,overlap.start_pos-1-seq.insert_start,"ActionInsert") # adds the ActionInsert to the sequence 
      a = seq.new_action(0,overlap.start_pos-1-seq.insert_start,"ActionInsert") # adds the ActionInsert to the sequence 
     
      actions.push a 
      # puts " 1---------- Inserto antes del linker en pos #{a.start_pos} #{a.end_pos}"
   elsif (overlap.start_pos>seq.insert_start)   #if overlap's positions are right and insert's size is short   
     # puts " 2---------- #{seq.insert_start},#{overlap.start_pos}-1-#{seq.insert_start}" 
     
        
     #It's created an ActionShortInsert before the ActionLinker
     # a = seq.new_action(seq.insert_start_last-seq.insert_start,overlap.start_pos-1-seq.insert_start,"ActionShortInsert") # adds the ActionInsert to the sequence 
     a = seq.new_action(0,overlap.start_pos-1-seq.insert_start,"ActionShortInsert") # adds the ActionInsert to the sequence 
     
     actions.push a  
     # puts " 2---------- Inserto corto antes del linker en pos #{a.start_pos} #{a.end_pos}" 
     
      
   end      
 
 
  @stats[:insert_size_left]={insert_size => 1}
  
end

#add_action_inserts(insert, linker, actions, seq) ⇒ Object

crea una action insert controlado si el inserto es corto o no y actualiza las stats según sea inserto izquierdo o derecho



32
33
34
35
36
37
38
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
70
71
72
73
74
75
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 32

def add_action_inserts(insert,linker,actions,seq) 

  insert_size = insert[1]-insert[0]+1
  
  min_insert_size = @params.get_param('min_insert_size_paired').to_i

  if  (insert_size >= min_insert_size)
      if (insert[0]<linker.start_pos) #insert is on the left from the linker 

        add_stats('left_insert_size',insert_size)
      elsif (insert[0]>linker.end_pos)   #insert is on the right from the linker 

        add_stats('right_insert_size',insert_size)         
      end  
      
      a = seq.new_action(insert[0]-seq.insert_start,insert[1]-seq.insert_start,"ActionInsert") # adds the ActionInsert to the sequence 
      actions.push a 

  else
     
    if (insert[0]<linker.start_pos) #insert is on the left from the linker 
      # @stats[:short_left_insert_size]={insert_size => 1}  
      add_stats('short_left_insert_size',insert_size)
    elsif (insert[0]>linker.end_pos)   #insert is on the right from the linker 
      # @stats[:short_right_insert_size]={insert_size => 1} 
      add_stats('short_right_insert_size',insert_size)
    end
    
    
    #create an ActionShortInsert before the ActionLinker

    # adds the ActionInsert to the sequence 
    a = seq.new_action(insert[0]-seq.insert_start,insert[1]-seq.insert_start,"ActionShortInsert") # adds the ActionInsert to the sequence 
    
 
    actions.push a  
      
   end      
 
 
     
   
  
end

#exec_seq(seq, blast_query) ⇒ Object 



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
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 279

def exec_seq(seq,blast_query)
   $LOG.debug "[#{self.class.to_s}, seq: #{seq.seq_name}]: extract inserts"    
        
   # puts "INSERTO ANTES LINKER INSERT:"+seq.seq_fasta

   #look for ActionLinker into the sequence's actions
   linkers=seq.get_actions(ActionLinker)       
   #look for ActionVectors into the sequence's actions
   vectors=seq.get_actions(ActionVectors)   
   #look for ActionLowQuality into the sequence's actions
   low_quals=seq.get_actions(ActionLowQuality)
   
   insert_size=0
   actions=[] 
   sub_inserts=[]
   
   if (linkers.count==1) #linker found   
      linker=linkers[0]
      
      # get left insert
      if linker.start_pos>seq.insert_start
        sub_inserts.push [ seq.insert_start,linker.start_pos-1 ]  
      end
      
		#get right insert
      if linker.end_pos<seq.insert_end
        sub_inserts.push [ linker.end_pos+1, seq.insert_end]
      end
      # puts '1ST SUBS:'
      # puts sub_inserts.join("\n")
		#split sub_inserts by vectors
      split_by(vectors,sub_inserts)
      # puts 'SUBS:'
      # puts sub_inserts.join("\n")
      
      #sub_inserts=select_the_best_with_linker(sub_inserts,linker) if not sub_inserts.empty?   
      
      # split by low qual actions
      split_by(low_quals,sub_inserts)
      # puts 'SUBS:'
      # puts sub_inserts.join("\n")
      sub_inserts=select_the_best_with_linker(sub_inserts,linker)  if not sub_inserts.empty?   
      
      if sub_inserts.empty?
        # if is an empty insert
        a=seq.new_action(0,0,'ActionEmptyInsert') 
        seq.seq_rejected=true 
        seq.seq_rejected_by_message='empty insert'
        actions.push a
      end
      
      sub_inserts.each do |sub_i| 
        add_action_inserts(sub_i,linker,actions,seq)  # ponerlo también abajo para que controle si la accion es de inserto corto o no
      end
      
    
   else # no linker found => add whole insert

     sub_inserts.push [ seq.insert_start, seq.insert_end ]
     
     split_by(vectors,sub_inserts)

     #sub_inserts=select_the_best(sub_inserts) if not sub_inserts.empty?   
     
     split_by(low_quals,sub_inserts)
     
     sub_inserts=select_the_best(sub_inserts) if not sub_inserts.empty?

     # ordena los subinsertos por tamaño 
     # sub_inserts.sort!{|i,j| j[1]-j[0]<=>i[1]-i[0]}
     
     if sub_inserts.empty?
       found_insert_size = 0  # position from an empty insert
       
       a=seq.new_action(0,0,'ActionEmptyInsert') # refactorizando codigo
       seq.seq_rejected=true
       seq.seq_rejected_by_message='empty insert'
     else
       found_insert_size =(sub_inserts[0][1]-sub_inserts[0][0]+1)
     end 


     if (found_insert_size >= (@params.get_param('min_insert_size_trimmed').to_i))     
       add_stats('insert_size',found_insert_size)
       a = seq.new_action(sub_inserts[0][0]-seq.insert_start, sub_inserts[0][1]-seq.insert_start,"ActionInsert") # adds the ActionInsert to the sequence before adding the actionMid   
     elsif (found_insert_size!=0)  # if is a short insert
       add_stats('short_insert_size',found_insert_size)                                                                                                
       a = seq.new_action(sub_inserts[0][0]-seq.insert_start, sub_inserts[0][1]-seq.insert_start,"ActionShortInsert") # adds the ActionInsert to the sequence before adding the actionMid   
       seq.seq_rejected=true 
       seq.seq_rejected_by_message='short insert'
     end
     actions.push a   
   end
   
   seq.add_actions(actions)    
                 
                 
   # find inserts to see if it is necessary to reject it
    
   if ! seq.seq_rejected
    inserts=seq.get_actions(ActionInsert)
    if inserts.empty? 
     seq.seq_rejected=true 
  
     if seq.get_actions(ActionShortInsert).empty?
       seq.seq_rejected_by_message='empty insert'
     else
	     seq.seq_rejected_by_message='short insert'
     end
    end     
   end
end

#part_left_overlap?(r1_start, r1_end, r2_start, r2_end) ⇒ Boolean

check if part of a vector is in a linker

Returns:

  • (Boolean) 


15
16
17
18
19
20
21
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 15

def part_left_overlap?(r1_start,r1_end,r2_start,r2_end)
  
   
   
   return ((r1_start<r2_start) and (r1_end<=r2_end)  and (r1_end>=r2_start) )  #overlap on the left of r2
           
end

#part_right_overlap?(r1_start, r1_end, r2_start, r2_end) ⇒ Boolean

Returns:

  • (Boolean) 


23
24
25
26
27
28
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 23

def part_right_overlap?(r1_start,r1_end,r2_start,r2_end)
 
   
   
   return ((r1_end>r2_end) and (r1_start<=r2_end)  and (r1_start>=r2_start) )    #overlap on the right of r2
end

#select_the_best(sub_inserts) ⇒ Object

select the best subinsert, when there is not a linker



217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 217

def select_the_best(sub_inserts)   
  
  insert_size = 0
     
  insert = nil
     
  sub_inserts.each do |sub_i|
    
    if (insert_size<(sub_i[1]-sub_i[0]+1))
      insert_size = (sub_i[1]-sub_i[0]+1) 
      insert=sub_i 
    end       
    
  end    
  
  sub_inserts=[]
  sub_inserts.push insert  if !insert.nil?
      
  # puts " subinsert #{sub_inserts.join(' ')}" 
  
  return sub_inserts
end

#select_the_best_with_linker(sub_inserts, linker) ⇒ Object

select the best subinsert when there is a linker



241
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
268
269
270
271
272
273
274
275
276
277
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 241

def select_the_best_with_linker(sub_inserts,linker)   
  
  left_insert_size = 0
  right_insert_size = 0 
  
  left_insert = nil
  right_insert = nil
 
  sub_inserts.each do |sub_i|
    #puts "*SBI: "+sub_i.join(',')
    if (sub_i[0]<linker.start_pos)   #if the subinsert is on the left from the linker 
      if (left_insert_size<(sub_i[1]-sub_i[0]+1))
        left_insert_size = (sub_i[1]-sub_i[0]+1) 
        left_insert=sub_i
        # puts " left"
      end       
      
    elsif (sub_i[0]>linker.end_pos)   #if the subinsert is on the right from the linker
      if (right_insert_size<(sub_i[1]-sub_i[0]+1))
        right_insert_size = (sub_i[1]-sub_i[0]+1) 
        right_insert=sub_i 
      end
        # puts " right"
    end
  end    
  # puts " left #{left_insert} #{left_insert_size} right #{right_insert} #{right_insert_size}"
  sub_inserts=[]
  sub_inserts.push left_insert  if !left_insert.nil?
  sub_inserts.push right_insert if !right_insert.nil?     
  # puts " subinsert #{sub_inserts.join(' ')}" 
  #puts "SELECTED SUBINSERTS"
#    sub_inserts.each do |sub_i|
#      puts "*SBI: "+sub_i.join(',')
#    end
#    
  return sub_inserts
end

#split_by(actions, sub_inserts) ⇒ Object 



163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
 
# File 'lib/seqtrimnext/plugins/plugin_extract_inserts.rb', line 163

def split_by(actions,sub_inserts) 
     
  delete=false 
  
  # puts " split #{sub_inserts.each{|i| i.join(',')}}"
  if !sub_inserts.empty?
    actions.each  do |action|  
      sub_inserts.reverse_each do |sub_i|
        # puts "A: [#{action.start_pos},#{action.end_pos}] cuts [#{sub_i[0]},#{sub_i[1]}] "
        if ((action.start_pos<=sub_i[0]) && (action.end_pos>=sub_i[1])) 
          # if not exists any subinsert
          delete=true
          
        elsif ((action.end_pos>=sub_i[0]) && (action.end_pos+1<=sub_i[1]))
          # if exists an subinsert  between the action one and the end of subinsert  
        
          sub_inserts.push [action.end_pos+1,sub_i[1]]        # mark subinsert after the action
        
          delete=true 
          # puts " !!!! 1" 
          if ((action.start_pos-1>=sub_i[0]))
            # if exists an subinsert  between the start of the subinsert and the action   
            sub_inserts.push [sub_i[0],action.start_pos-1]    # mark subinsert before the action
            delete=true

            # puts " !!!! 2-1" 
          end 
        
        elsif ((action.start_pos-1>=sub_i[0]) && (action.start_pos<=sub_i[1]))
            # if exists an subinsert  between the start of the subinsert and the action   
            sub_inserts.push [sub_i[0],action.start_pos-1,]   # mark subinsert before the action 
            delete=true

            # puts " !!!! 2-2" 
        
        
        end  
      
      
        # sub_inserts.delete [sub_i[0],sub_i[1]] and delete=false  and puts " DELETEEE ___________  #{delete}" if delete
        if delete  
          sub_inserts.delete [sub_i[0],sub_i[1]] 
          delete=false  
          # puts " DELETEEE ___________  #{delete} #{[sub_i[0] , sub_i[1]] }"     
        end
        # puts " eee #{sub_inserts.join(',')}" 
             
                                                                     
      end #each sub_insert
    end #each action
  end
end