Parse prodigal gene prediction output

convert_prodigal_to_yml.rb

require 'rubygems'
require 'bio'
require 'yaml'

locations = Hash.new{|k,v| k[v] = []}
Bio::FlatFile.open(Bio::GenBank, ARGV[0]).each do |sequence|
  sequence.each_cds do |cds|
    cds.locations.each do |location|
      locations[sequence.definition] << {
        'start'      => location.from,
        'end'        => location.to,
        'complement' => location.strand < 0}
    end
  end
end

puts YAML.dump(locations)

map_yml_to_fasta.rb

require 'rubygems'
require 'bio'
require 'yaml'

coordinates = YAML.load(File.read(ARGV[0]))
entries = Bio::FastaFormat.open(ARGV[1])

entries.each do |entry|
  # Had to normalise white space in the name
  name = entry.definition.gsub(/\s+/,' ')
  coordinates[name].each_with_index do |coords,i|
    seq = entry.seq.subseq(coords['start'],coords['end'])
    seq = seq).complement if coords['complement']
    puts seq.to_fasta("#{name} ORF:#{i + 1}")
  end
end