Module: Bio::PolyploidTools::Mask

Defined in:
lib/bio/PolyploidTools/Mask.rb

Class Method Summary collapse

Class Method Details

.find_end(seqs) ⇒ Object



14
15
16
17
18
19
20
21
22
23
24
# File 'lib/bio/PolyploidTools/Mask.rb', line 14

def self.find_end(seqs)
  size = seqs.values[0].size
  names = seqs.keys
  i = size - 1
  gap_count = 3
  while i > 0 and gap_count > 0
    gap_count = names.map { |chr| seqs[chr][i] == "-" ? 1:0  }.inject(0, :+)
    i -= 1
  end
  i + 1
end

.find_start(seqs) ⇒ Object



26
27
28
29
30
31
32
33
34
35
36
37
# File 'lib/bio/PolyploidTools/Mask.rb', line 26

def self.find_start(seqs)
  size = seqs.values[0].size
  names = seqs.keys
  i = 0
  gap_count = 3
  while i < size  and gap_count > 0
    gap_count = names.map { |chr| seqs[chr][i] == "-" ? 1 : 0  } .inject(0, :+)

    i += 1
  end
  i - 1
end

.get(seqs, target: nil, seq_start: 0, seq_end: 0) ⇒ Object



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
# File 'lib/bio/PolyploidTools/Mask.rb', line 39

def self.get(seqs, target: nil, seq_start: 0, seq_end: 0)
  names = seqs.keys
  target = names[0] if target.nil?
  masked_snps = seqs[target].downcase
  i = 0
  while i < masked_snps.size
    different = 0
    cov = 0
    gap = false
    names.each do | chr |
      if seqs[chr][i]  != "-" and seqs[chr][i]  != "n" and seqs[chr][i]  != "N"
        cov += 1
      end
      if chr != target
       different += 1  if masked_snps[i].upcase != seqs[chr][i].upcase
      end
      if seqs[chr][i]  == "-" and chr == target
          gap = true
      end
    end
    masked_snps[i] = "." if different == 0
    masked_snps[i] = "." if cov == 1
    masked_snps[i] = "*" if cov == 0
    expected_snps  = names.size - 1
    masked_snps[i] = masked_snps[i].upcase if different == expected_snps
    if gap
      masked_snps[i] = different == expected_snps ? "-" : "_"
    end
    masked_snps[i] = "|" if i < seq_start or i > seq_end
    i += 1
  end
  masked_snps
end

.stats(mask, triad, gene, genome, reference) ⇒ Object



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
# File 'lib/bio/PolyploidTools/Mask.rb', line 73

def self.stats(mask, triad, gene, genome, reference)
  specific = []
  semispecific = []
  sp_i = 0
  semi = 0
  i = 0
  mask.to_s.each_char do |e|
    case e
    when "n","N"
      i += 1
    when /[[:lower:]]/ then
      semispecific << semi
      semi = 0
      i += 1
    when /[[:upper:]]/ then
      specific     << sp_i
      semispecific << semi
      sp_i = 0
      semi = 0
      i += 1
    when "." then
      semi += 1
      sp_i += 1
      i += 1
    end
  end
  {
    reference: reference,
    triad: triad,
    genome: genome,
    gene: gene,
    semispecific_mean: semispecific.mean,
    semispecific_bases: semispecific.size,
    semispecific_identity: (1 - (semispecific.size.to_f / i)) * 100 ,
    specific_mean: specific.mean,
    specific_bases: specific.size,
    specific_identity: (1 - (specific.size.to_f / i )) * 100,
    aligned_length: i,
    specific: specific,
    semispecific: semispecific
  }
end