Class: PluginAmplicons

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

Overview

Author: Almudena Bocinos Rioboo

Defines the main methods that are necessary to execute PluginAdapters Inherit: Plugin

Instance Attribute Summary

Attributes inherited from Plugin

#stats

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from Plugin

#add_stats, #add_text_stats, auto_setup, check_param, get_graph_filename, get_graph_title, graph_ignored?, ignored_graphs, #initialize, #merge_hits, #overlapX?, plot_setup, 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



147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
# File 'lib/seqtrimnext/plugins/plugin_amplicons.rb', line 147

def self.check_params(params)
  errors=[]

  comment='Blast E-value used as cut-off when searching for primers'
  # default_value = 1e-6
  default_value = 1
  params.check_param(errors,'blast_evalue_primers','Float',default_value,comment)

  comment='Minimum required identity (%) for a reliable primer'
  default_value = 95
  params.check_param(errors,'blast_percent_primers','Integer',default_value,comment)

  comment='Minimun primer size'
  default_value = 15
  params.check_param(errors,'min_primer_size','Integer',default_value,comment)

  comment='Path for primers database'
  # default_value = File.join($FORMATTED_DB_PATH,'adapters_ab.fasta')
  default_value=nil
  params.check_param(errors,'primers_db','DB',default_value,comment)

  return errors
end

Instance Method Details

#do_blasts(seqs) ⇒ Object



25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
# File 'lib/seqtrimnext/plugins/plugin_amplicons.rb', line 25

def do_blasts(seqs)
  # find MIDS  with less results than max_target_seqs value
  blast=BatchBlast.new("-db #{@params.get_param('primers_db')}",'blastn'," -task blastn-short -perc_identity #{@params.get_param('blast_percent_primers')}")
  $LOG.debug('BLAST:'+blast.get_blast_cmd)

  fastas=[]

  seqs.each do |seq|
    fastas.push ">"+seq.seq_name
    fastas.push seq.seq_fasta
  end

  # fastas=fastas.join("\n")

  blast_table_results = blast.do_blast(fastas)

  # puts blast_table_results.inspect

  return blast_table_results
end

#exec_seq(seq, blast_query) ⇒ Object



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
76
77
78
79
80
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
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
# File 'lib/seqtrimnext/plugins/plugin_amplicons.rb', line 47

def exec_seq(seq,blast_query)
  if blast_query.query_id != seq.seq_name
    raise "Blast and seq names does not match, blast:#{blast_query.query_id} sn:#{seq.seq_name}"
  end

  $LOG.debug "[#{self.class.to_s}, seq: #{seq.seq_name}]: looking for primers into the sequence"

  # puts blast_query.inspect
  
  # merge hits
  # primers=blast_query.merged_hits!
  
  # do not merge hits, since id is important
  primers=blast_query.hits
  
  min_primer_size=@params.get_param('min_primer_size').to_i
  # puts "MERGED:"
  # puts primers.inspect

  # type = 'ActionAbAdapter'
  actions=[]
  adapter_size=0

  # filter primers by size
  primers = primers.select{|primer| (primer.size >= min_primer_size)}.sort{|p1,p2| p1.size<=>p2.size}.reverse
  # puts "FILTERED:"
  # puts primers.inspect
  
  # reject sequences with less than two primers
  if primers.count < 2

    seq.seq_rejected=true
    seq.seq_rejected_by_message='Primer pair not found'

    # @stats[:rejected_seqs]={'rejected_seqs_by_contaminants' => 1}
    add_stats('rejected','primers_not_found')

  else # has two primers, or more

    if seq.seq_fasta.index('N')
      seq.seq_rejected=true
      seq.seq_rejected_by_message='At least one N found'

      # @stats[:rejected_seqs]={'rejected_seqs_by_contaminants' => 1}
      add_stats('rejected','one_n_found')

    else
      # puts "EL DE ARRIBA"
      
      # take first two primers and order them by qbeg
      left_primer = primers[0..1].sort{|p1,p2| p1.q_beg<=>p2.q_beg}.first
      right_primer = primers[0..1].sort{|p1,p2| p1.q_beg<=>p2.q_beg}.last
      
      # puts "LEFT_PRIMER:"
      # puts left_primer.inspect
      # puts "RIGHT_PRIMER:"
      # puts right_primer.inspect

      # if (left_primer.size>= min_primer_size) && (right_primer.size>= min_primer_size)

        a = seq.new_action(left_primer.q_beg,left_primer.q_end,'ActionLeftPrimer')
        a.message = left_primer.subject_id
        a.tag_id = left_primer.subject_id
        a.reversed = left_primer.reversed
        a.left_action = true
        actions.push a

        add_stats('primer_size',left_primer.size)
        add_stats('primer_id',left_primer.subject_id)

        a = seq.new_action(right_primer.q_beg,right_primer.q_end,'ActionRightPrimer')
        a.message = right_primer.subject_id
        a.reversed = right_primer.reversed
        a.tag_id = right_primer.subject_id
        a.right_action = true
        actions.push a

        add_stats('primer_size',right_primer.size)
        add_stats('primer_id',right_primer.subject_id)
        
        seq.add_file_tag(2, left_primer.subject_id, :file)
        seq.add_file_tag(2, right_primer.subject_id, :file)
        

      # end


      if !actions.empty?
        seq.add_actions(actions)
        add_stats('sequences_with_primers','count')
        
        # add_stats('sequences',seq.seq_fasta)
      end

    end
    #
  end
end

#execute(seqs) ⇒ Object

adapters found at end of sequence are even 2 nt wide, cut in 5 because of statistics MIN_PRIMER_SIZE = 5 MIN_FAR_ADAPTER_SIZE = 13 MIN_LEFT_ADAPTER_SIZE = 9 Begins the plugin1’s execution to warn that there are contaminants in the sequence “seq”



17
18
19
20
21
22
23
# File 'lib/seqtrimnext/plugins/plugin_amplicons.rb', line 17

def execute(seqs)
  blasts= do_blasts(seqs)

  seqs.each_with_index do |s,i|
    exec_seq(s,blasts.querys[i])
  end
end