Method: Bio::Alignment::Output#output_molphy

Defined in:
lib/bio/alignment.rb

#output_molphy(options = {}) ⇒ Object

Generates Molphy alignment format text as a string



1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
 
# File 'lib/bio/alignment.rb', line 1151

def output_molphy(options = {})
  len = self.alignment_length
  header = "#{self.number_of_sequences} #{len}\n"
  sn = self.sequence_names.collect { |x| x.to_s.gsub(/[\r\n\x00]/, ' ') }
  if options[:replace_space]
    sn.collect! { |x| x.gsub(/\s/, '_') }
  end
  if !options.has_key?(:escape) or options[:escape]
    sn.collect! { |x| x.gsub(/[\:\;\,\(\)]/, '_') }
  end
  if !options.has_key?(:split) or options[:split]
    sn.collect! { |x| x.split(/\s/)[0].to_s }
  end
  if !options.has_key?(:avoid_same_name) or options[:avoid_same_name]
    sn = __clustal_avoid_same_name(sn, 30)
  end

  seqwidth  = (options[:width] or 60)
  seqregexp = Regexp.new("(.{1,#{seqwidth}})")
  gchar = (options[:gap_char] or '-')

  aseqs = Array.new(len).clear
  self.each_seq do |s|
    aseqs << s.to_s.gsub(self.gap_regexp, gchar)
  end
  case options[:case].to_s
  when /lower/i
    aseqs.each { |s| s.downcase! }
  when /upper/i
    aseqs.each { |s| s.upcase! }
  end
  
  aseqs.collect! do |s|
    s << (gchar * (len - s.length))
    s.gsub!(seqregexp, "\\1\n")
    sn.shift + "\n" + s
  end
  aseqs.unshift(header)
  aseqs.join('')
end