Class: FastaManip

Inherits:
Object
  • Object
show all
Defined in:
lib/bacterial-annotator/fasta-manip.rb

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(fasta_file, meta) ⇒ FastaManip

Initialize fasta holder



16
17
18
19
20
21
22
23
24
25
26
27
28
29
# File 'lib/bacterial-annotator/fasta-manip.rb', line 16

def initialize fasta_file, meta

  @fasta_file = fasta_file
  @fasta_flat = Bio::FlatFile.auto(@fasta_file)
  @meta = meta
  @prodigal_files = nil
  @single_fasta = nil
  @seq_info = nil

  if @fasta_flat.dbclass != Bio::FastaFormat
    abort "Aborting : The input sequence is not a fasta file !"
  end

end

Instance Attribute Details

#fasta_fileObject (readonly)

Returns the value of attribute fasta_file.



13
14
15
# File 'lib/bacterial-annotator/fasta-manip.rb', line 13

def fasta_file
  @fasta_file
end

#fasta_flatObject (readonly)

Returns the value of attribute fasta_flat.



13
14
15
# File 'lib/bacterial-annotator/fasta-manip.rb', line 13

def fasta_flat
  @fasta_flat
end

#prodigal_filesObject (readonly)

Returns the value of attribute prodigal_files.



13
14
15
# File 'lib/bacterial-annotator/fasta-manip.rb', line 13

def prodigal_files
  @prodigal_files
end

Instance Method Details

Utility function to print the sequence to the end of a gbk file



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
# File 'lib/bacterial-annotator/fasta-manip.rb', line 106

def print_sequence_for_gbk seq

  outseq = "ORIGIN\n"
  # puts "ORIGIN"

  ntNum = 0
  sequence = seq.seq.downcase

  nt_left = true
  it = 0

  while nt_left

    if sequence.length > it+60
      nt_to_add = sequence[it..(it+59)]
      # printf "%9s ", (ntNum - l.size + 2)
      outseq += "%9s " % (it+1)
      outseq += nt_to_add.scan(/.{1,10}/).join(" ")
      outseq += "\n"
      it += 60
    else
      nt_to_add = sequence[it..sequence.length-1]
      outseq += "%9s " % (it+1)
      outseq += nt_to_add.scan(/.{1,10}/).join(" ")
      outseq += "\n"
      outseq += "//"
      nt_left = false
    end

  end

  return outseq, sequence.length

end

#run_prodigal(root, outdir) ⇒ Object

Run prodigal on the genome to annotate



32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
# File 'lib/bacterial-annotator/fasta-manip.rb', line 32

def run_prodigal root, outdir
  @prodigal_files = {}
  Dir.mkdir "#{outdir}" if ! Dir.exists? "#{outdir}"
  if @meta
    system("#{root}/prodigal.linux -p meta -i #{@fasta_file} -a #{outdir}/Proteins.fa -d #{outdir}/Genes.fa -o #{outdir}/Genbanks.gbk -q")
  else
    system("#{root}/prodigal.linux -i #{@fasta_file} -a #{outdir}/Proteins.fa -d #{outdir}/Genes.fa -o #{outdir}/Genbanks.gbk -q")
  end

  @prodigal_files = {multiGBK: "#{outdir}/Genbanks.gbk",
                     contigs: [],
                     contigs_length: [],
                     genes: "#{outdir}/Genes.fa",
                     proteins: "#{outdir}/Proteins.fa",
                     prot_ids_by_contig: {},
                     fasta_path: "#{outdir}/single-fasta/",
                     gbk_path: "#{outdir}/single-genbank/"}
  split_fasta outdir
  split_genbank outdir, "#{outdir}/Genbanks.gbk"
  extract_cds_names
  @prodigal_files
end

#split_fasta(outdir) ⇒ Object

Split Multi Genbanks file RETURN : array of fasta files



58
59
60
61
62
63
64
65
66
67
68
69
70
# File 'lib/bacterial-annotator/fasta-manip.rb', line 58

def split_fasta outdir
  @single_fasta = {}
  Dir.mkdir("#{outdir}/single-fasta") if ! Dir.exists?("#{outdir}/single-fasta")
  @fasta_flat.each_entry do |seq|
    file_name = seq.definition.chomp.split(" ")[0]
    @prodigal_files[:contigs] << "#{file_name}"
    @prodigal_files[:contigs_length] << seq.seq.length
    File.open("#{outdir}/single-fasta/#{file_name}.fasta", "w") do |fwrite|
      fwrite.write(seq)
    end
    @single_fasta[file_name] = seq
  end
end

#split_genbank(outdir, multigbk) ⇒ Object

Split Multi Genbanks file RETURN : array of genbank files



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
# File 'lib/bacterial-annotator/fasta-manip.rb', line 75

def split_genbank outdir, multigbk

  Dir.mkdir("#{outdir}/single-genbank")if ! Dir.exists?("#{outdir}/single-genbank")
  File.open(multigbk,"r") do |f|
    fopen = nil
    while l = f.gets
      if l[0..9] == "DEFINITION"
        file_name = l.chomp.split(";")[2].gsub("seqhdr","").delete("\"").delete("=").split(" ")[0]
        outseq, seq_length = print_sequence_for_gbk @single_fasta[file_name]
        spacer = " " * (20-seq_length.to_s.length)
        date = DateTime.now
        month = Date::ABBR_MONTHNAMES[date.month]
        day = "%02d" % date.day
        year = date.year
        locus = "LOCUS       #{file_name}#{spacer}#{seq_length.to_s} bp    DNA     linear   BCT #{day}-#{month}-#{year}\n"
        locus += "DEFINITION  #{file_name}\n"
        fopen = File.open("#{outdir}/single-genbank/#{file_name}.gbk", "w")
        fopen.write(locus)
      elsif l[0..1] == "//"
        fopen.write(outseq)
        fopen.close
      elsif ! l.include? "                     /note="
        fopen.write(l)
      end
    end
  end

end