Class: NpSearch::Validators

Inherits:
Object
  • Object
show all
Defined in:
lib/npsearch/arg_validator.rb

Instance Method Summary collapse

Instance Method Details

#motif_type(motif) ⇒ Object

Guesses the type of the data in the supplied motif. It ignores all

non-word characters (e.g. '|' that is used for regex). It has a 90%
threshold.



254
255
256
257
258
259
260
261
262
 
# File 'lib/npsearch/arg_validator.rb', line 254

def motif_type(motif)
  motif_seq = Bio::Sequence.new(motif.gsub(/\W/, ''))
  type = motif_seq.guess(0.9)
  return unless type.to_s != 'Bio::Sequence::AA'
  fail IOError('Critical Error: There seems to be an error in' \
                ' processing the motif. Please ensure that the motif' \
                ' contains amino acid residues that you wish to search' \
                ' for.')
end

#output_dir(output_dir) ⇒ Object

Checks for the presence of the output directory; if not found, it asks

the user whether they want to create the output directory.



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
 
# File 'lib/npsearch/arg_validator.rb', line 111

def output_dir(output_dir)
  unless File.directory? output_dir # If output_dir doesn't exist
    fail IOError, "\n\nThe output directory deoes not exist\n\n"
  end
rescue IOError
  puts # a blank line
  puts 'The output directory does not exist.'
  puts # a blank line
  puts "The directory '#{output_dir}' will be created in this location."
  puts 'Do you to continue? [y/n]'
  print '> '
  inp = $stdin.gets.chomp
  until inp.downcase == 'n' || inp.downcase == 'y' || inp == ''
    puts # a blank line
    puts "The input: '#{inp}' is not recognised - 'y' or 'n' are the" \
         ' only recognisable inputs.'
    puts 'Please try again.'
    puts "The directory '#{output_dir}' will be created in this" \
         ' location.'
    puts 'Do you to continue? [y/n]'
    print '> '
    inp = $stdin.gets.chomp
  end
  if inp.downcase == 'y' || inp == ''
    FileUtils.mkdir_p "#{output_dir}"
    puts 'Created output directory...'
  elsif inp.downcase == 'n'
    raise ArgumentError('Critical Error: An output directory is' \
                        ' required; please create an output directory' \
                        ' and then try again.')
  end
end

#signalp_dir ⇒ Object

Ensures that the Signal P Script is present. If not found in the home

directory, it asks the user for its location.



146
147
148
149
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
 
# File 'lib/npsearch/arg_validator.rb', line 146

def signalp_dir
  signalp_dir = "#{Dir.home}/SignalPeptide"
  if File.exist? "#{signalp_dir}/signalp"
    signalp_directory = signalp_dir
  else
    begin
      fail IOError('The Signal P Script directory cannot be found at' \
                    " the following location: '#{signalp_dir}/'.")
    rescue IOError
      puts # a blank line
      puts 'Error: The Signal P Script directory cannot be found at the' \
           " following location: '#{signalp_dir}/'."
      puts # a blank line
      puts 'Please enter the full path or a relative path to the Signal' \
           ' P Script directory (i.e. to the folder containing the' \
           ' Signal P script). Refer to the online tutorial for more help'
      print '> '
      inp = $stdin.gets.chomp
      until (File.exist? "#{signalp_dir}/signalp") ||
            (File.exist? "#{inp}/signalp")
        puts # a blank line
        puts 'The Signal P directory cannot be found at the following' \
             " location: '#{inp}'"
        puts 'Please enter the full path or a relative path to the Signal' \
             ' Peptide directory again.'
        print '> '
        inp = $stdin.gets.chomp
      end
      signalp_directory = inp
      puts # a blank line
      puts "The Signal P directory has been found at '#{signalp_directory}'"
      FileUtils.ln_s "#{signalp_directory}", "#{Dir.home}/SignalPeptide",
                     force: true
      puts # a blank line
    end
  end
  signalp_directory
end

#sp_column(_input_file) ⇒ Object

Ensures that the critical columns in the tabular results produced by the

Signal P script are conserved. Run from the 'sp_results' method.



200
201
202
203
204
205
206
207
208
209
210
211
 
# File 'lib/npsearch/arg_validator.rb', line 200

def sp_column(_input_file)
  File.open('signalp_out.txt', 'r') do |file_stream|
    secondline = file_stream.readlines[1]
    row = secondline.gsub(/\s+/m, ' ').chomp.split(' ')
    if row[1] != 'name' && row[4] != 'Ymax' && row[5] != 'pos' &&
       row[9] != 'D'
      return true
    else
      return false
    end
  end
end

#sp_results(signalp_output_file) ⇒ Object

Ensure that the right version of the Signal P script is used (via

'sp_version' Method). If the wrong signal p script has been linked to
NpSearch, check whether the critical columns in the tabular results
produced by the Signal P Script are conserved (via 'sp_column'
Method).



218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
 
# File 'lib/npsearch/arg_validator.rb', line 218

def sp_results(signalp_output_file)
  return if sp_version(signalp_output_file)
  # i.e. if Signal P is the wrong version
  if sp_column(signalp_output_file) # If wrong version but correct columns
    puts # a blank line
    puts 'Warning: The wrong version of signalp has been linked.' \
         ' However, the signal peptide output file still seems to' \
         ' be in the right format.'
  else
    puts # a blank line
    puts 'Warning: The wrong version of the signal p has been linked' \
         ' and the signal peptide output is in an unrecognised format.'
    puts 'Continuing may give you meaningless results.'
  end
  puts # a blank line
  puts 'Do you still want to continue? [y/n]'
  print '> '
  inp = $stdin.gets.chomp
  until inp.downcase == 'n' || inp.downcase == 'y'
    puts # a blank line
    puts "The input: '#{inp}' is not recognised - 'y' or 'n' are the" \
         ' only recognisable inputs.'
    puts 'Please try again.'
  end
  if inp.downcase == 'y'
    puts 'Continuing.'
  elsif inp.downcase == 'n'
    fail IOError('Critical Error: NpSearch only supports SignalP 4.1' \
                  ' (downloadable form CBS) Please ensure the version' \
                  ' of the signal p script is downloaded.')
  end
end

#sp_version(input_file) ⇒ Object

Ensures that the supported version of the Signal P Script has been linked

to NpSearch. Run from the 'sp_results' method.



187
188
189
190
191
192
193
194
195
196
 
# File 'lib/npsearch/arg_validator.rb', line 187

def sp_version(input_file)
  File.open(input_file, 'r') do |file_stream|
    first_line = file_stream.readline
    if first_line.match(/# SignalP-4.1/)
      return true
    else
      return false
    end
  end
end