bioinform 0.1.7 → 0.1.8

data/TODO.txt

@@ -1,11 +1,14 @@
 ToDo:
   how to make PM#equal? and PM#hash so that using PMs in Sets wouldn't destroy comparability of Sets and two sets with the same PMs(but different objects) would be equal. (also using pm as a hash-key)
-  
+  Make specs and fix code in such a way that Parser.split_on_motifs and so on returned consistent result. E.g. Parser.parse! raised an error on multiple times invocation
+
+  refactor CLI::SplitMotifs in place where it splits collection file and choose real data models or makes PM
+
 Make parser exception print out text where parsing was broken (processing line +- 2 nearest lines and command and line numbers)
 Prevent parser going into infinity loop
 
 Create CLI-apps:
--- to merge many files(or whole folder) to a Collection
+-- to merge many files(or whole folder) to a Collection (in a way that makes able to give collection a name)
 
 Make Parsers to be switcheable in runtime so that one could parse string composed of two motifs in different formats.
 
@@ -18,6 +21,8 @@ Decide:
 -- should background be in PM by default?
 -- refactor PM.new #== and so on to make possible consistently introduce or remove a variable at a single line
 -- Make PCM#valid? and PPM#valid? more specific. This shouldn't destroy functionality to load arbitrary data as matrix, but only in force mode (I don't understand yet where should it be: in a constructor or where? And which validation-"severity" levels should be? Strong validation - size-only-validation - size-and-type-validation - no validation ???  or may be  options: valid_strictness: 'strict', 'usual', 'strict_with_name' ??? It should be considered)
+-- PM#to_pcm and friends have unintuitive behavior. E.g. pm.to_pcm.to_pwm != pm.to_pwm  First is matrix treated as pcm and then converted, while second is matrix treated as pwm from start
+-- Should parser be reloadable or not? May be delete #reset_scanner?
 
 Specs
 -- PWM#probabilities, #score_variance, #gauss_estimation

data/bin/merge_into_collection

@@ -0,0 +1,4 @@
+#!/usr/bin/env ruby
+
+require_relative '../lib/bioinform/cli/merge_into_collection'
+Bioinform::CLI::MergeIntoCollection.main(ARGV)

data/bin/pcm2pwm

@@ -1,4 +1,4 @@
 #!/usr/bin/env ruby
 
-require 'bioinform/cli/pcm2pwm'
+require_relative '../lib/bioinform/cli/pcm2pwm'
 Bioinform::CLI::PCM2PWM.main(ARGV)

data/bin/split_motifs

@@ -1,4 +1,4 @@
 #!/usr/bin/env ruby
 
-require 'bioinform/cli/split_motifs'
+require_relative '../lib/bioinform/cli/split_motifs'
 Bioinform::CLI::SplitMotifs.main(ARGV)

data/bioinform.gemspec

@@ -16,6 +16,8 @@ Gem::Specification.new do |gem|
   gem.version       = Bioinform::VERSION
   
   gem.add_dependency('activesupport', '>= 3.0.0')
+  gem.add_dependency('docopt', '>= 0.5.0')
   
   gem.add_development_dependency "rspec", ">= 2.0"
+  gem.add_development_dependency "fabrication", ">= 2.2.3"
 end

data/lib/bioinform/cli/merge_into_collection.rb

@@ -0,0 +1,76 @@
+require_relative '../../bioinform'
+require 'docopt'
+require 'shellwords'
+require 'yaml'
+
+module Bioinform
+  module CLI
+    module MergeIntoCollection
+      extend Bioinform::CLI::Helpers
+      def self.main(argv)
+        doc = <<-DOCOPT
+          Tool for merging multiple motifs into a single collection file.
+          It takes motif files or (entire collections) and creates a collection consisting of them all. By default motifs are treated simply as matrices(PM), but you can (possibly should) specify data model. Output file by default are in YAML-format but it's possible to create plain text file. YAML collections are useful if you want to provide additional information for motifs in collection with another tool, plain text is more readable by humans.
+
+          Usage:
+            #{__FILE__} [options] [<pm-files>...]
+
+          Options:
+            -h --help                 Show this screen.
+            -n --name NAME            Specify name for a collection. Default filename is based on this parameter
+            -o --output-file FILE     Output file for resulting collection
+            -m --data-model MODEL     Data model: PM, PCM, PPM or PWM [default: PM]
+            -p --plain-text           Output collection of motifs in plain text (motifs separated with newlines, no additional information included).
+        DOCOPT
+
+        doc.gsub!(/^#{doc[/\A +/]}/,'')
+        options = Docopt::docopt(doc, argv: argv)
+
+        plain_text = options['--plain-text']
+        name = options['--name']
+        if options['--plain-text']
+          output_file = options['--output-file'] || set_extension(name || 'collection', 'txt')
+        else
+          output_file = options['--output-file'] || set_extension(name || 'collection', 'yaml')
+        end
+        data_model = Bioinform.const_get(options['--data-model'].upcase)
+
+        if options['<pm-files>'].empty?
+          filelist = $stdin.read.shellsplit
+        else
+          filelist = options['<pm-files>']
+        end
+        
+        filelist = filelist.map do |data_source|
+          if File.directory? data_source
+            Dir.glob(File.join(data_source, '*'))
+          elsif File.file? data_source
+            data_source
+          else
+            raise "File or directory #{data_source} can't be found"
+          end
+        end.flatten
+
+        collection = Collection.new
+        collection.name = name  if name
+
+        filelist.each do |filename|
+          data_model.split_on_motifs(File.read(filename)).each do |pm|
+            pm.name ||= File.basename(filename, File.extname(filename))
+            collection << pm
+          end
+        end
+
+        if plain_text
+          File.open(output_file, 'w'){|f| f.puts(collection.to_s(false)) }
+        else
+          File.open(output_file, 'w'){|f| YAML.dump(collection, f) }
+        end
+        
+      rescue Docopt::Exit => e
+        puts e.message
+      end
+
+    end
+  end
+end

data/lib/bioinform/cli/pcm2pwm.rb

@@ -1,39 +1,39 @@
-require 'bioinform'
+require_relative '../../bioinform'
 require 'docopt'
 require 'shellwords'
 
 module Bioinform
-  module CLI
+  module CLI  
     module PCM2PWM
+      extend Bioinform::CLI::Helpers
       def self.main(argv)
         doc = <<-DOCOPT
-PCM to PWM converter.
-It transforms files with PCMs into files with PWMs. Folder for resulting files to save files can be specified. Resulting PWM files have the same name as original file but have another extension (.pwm by default).
-When filelist is empty, it's obtained from STDIN. One can use it: `ls -b pcm_folder/*.pcm | pcm2pwm` (ls -b option escape spaces in filenames)
+          PCM to PWM converter.
+          It transforms files with PCMs into files with PWMs. Folder for resulting files to save files can be specified. Resulting PWM files have the same name as original file but have another extension (.pwm by default).
+          When filelist is empty, it's obtained from STDIN. One can use it: `ls -b pcm_folder/*.pcm | pcm2pwm` (ls -b option escape spaces in filenames)
 
-Usage:
-  #{__FILE__} [options] [<pcm-files>...]
+          Usage:
+            #{__FILE__} [options] [<pcm-files>...]
 
-Options:
-  -h --help           Show this screen.
-  -e --extension EXT  Extension of output files [default: pwm]
-  -f --folder FOLDER  Where to save output files [default: .]
+          Options:
+            -h --help           Show this screen.
+            -e --extension EXT  Extension of output files [default: pwm]
+            -f --folder FOLDER  Where to save output files [default: .]
         DOCOPT
 
+        doc.gsub!(/^#{doc[/\A +/]}/,'')
         options = Docopt::docopt(doc, argv: argv)
 
-        if options['<pcm-files>'].empty?
-          filelist = $stdin.read.shellsplit
-        else
-          filelist = options['<pcm-files>']
-        end
-
+        pcm_files = options['<pcm-files>']
         folder = options['--folder']
+        extension = options['--extension']
+
         Dir.mkdir(folder)  unless Dir.exist?(folder)
+        filelist = (pcm_files.empty?)  ?  $stdin.read.shellsplit  :  pcm_files
 
-        filelist.each do |pcm_filename|
-          pwm = Bioinform::PCM.new( File.read(pcm_filename) ).to_pwm
-          File.open(Bioinform::CLI.output_filename(pcm_filename, options['--extension'], folder), 'w') do |f|
+        filelist.each do |filename|
+          pwm = Bioinform::PCM.new( File.read(filename) ).to_pwm
+          File.open(change_folder_and_extension(filename, extension, folder), 'w') do |f|
             f.puts pwm
           end
         end

data/lib/bioinform/cli/split_motifs.rb

@@ -1,40 +1,41 @@
-require 'bioinform'
+require_relative '../../bioinform'
 require 'docopt'
 
 module Bioinform
   module CLI
     module SplitMotifs
-
+      extend Bioinform::CLI::Helpers
       def self.main(argv)
         doc = <<-DOCOPT
-Motif splitter.
-It get a file with a set of motifs and splits it into motifs according to their names.
+          Motif splitter.
+          It get a file with a set of motifs and splits it into motifs according to their names.
 
-Usage:
-  #{__FILE__} [options] <collection-file>
+          Usage:
+            #{__FILE__} [options] <collection-file>
 
-Options:
-  -h --help                Show this screen.
-  -m --data-model MODEL    Data model: PM, PCM, PPM or PWM [default: PM]
-  -e --extension EXT       Extension of output files (by default it's based on data model)
-  -f --folder FOLDER       Where to save output files [default: .]
+          Options:
+            -h --help                Show this screen.
+            -e --extension EXT       Extension of output files
+            -f --folder FOLDER       Where to save output files [default: .]
         DOCOPT
 
+        doc.gsub!(/^#{doc[/\A +/]}/,'')
         options = Docopt::docopt(doc, argv: argv)
 
         folder = options['--folder']
-        Dir.mkdir(folder)  unless Dir.exist?(folder)
-
-        data_model = Bioinform.const_get(options['--data-model'].upcase)
-        extension = options['--extension'] || options['--data-model'].downcase
-
+        extension = options['--extension']
         collection_filename = options['<collection-file>']
+
+        Dir.mkdir(folder)  unless Dir.exist?(folder)
         raise "File #{collection_filename} not exist"  unless File.exist? collection_filename
-        input = File.read(collection_filename)
 
-        data_model.choose_parser(input).split_on_motifs(input, data_model).each do |motif|
-          File.open(File.join(folder, "#{motif.name}.#{extension}"), 'w') do |f|
-            f.puts motif
+        input = File.read(collection_filename)
+        Parser.choose(input).split.each do |motif|
+          if motif.is_a? PM
+            File.open(set_folder(folder, set_extension(motif.name, extension || motif.class.name.gsub(/^.*::/,'').downcase)), 'w'){|f| f.puts motif}
+          else
+            motif = PM.new(motif)
+            File.open(set_folder(folder, set_extension(motif.name, extension || 'mat')), 'w'){|f| f.puts motif}
           end
         end
       rescue Docopt::Exit => e

data/lib/bioinform/cli.rb

@@ -1,7 +1,21 @@
 module Bioinform
   module CLI
-    def self.output_filename(input_filename, extension, folder)
-      File.join(folder, File.basename(input_filename, File.extname(input_filename)) + ".#{extension}")
+    module Helpers
+      def basename_wo_extension(filename)
+        File.basename(filename, File.extname(filename))
+      end
+      def set_extension(filename, extension)
+        "#{filename}.#{extension}"
+      end
+      def set_folder(folder, filename)
+        File.join(folder, filename)
+      end
+      def change_extension(filename, extension)
+        set_extension(basename_wo_extension(filename), extension)
+      end
+      def change_folder_and_extension(input_filename, extension, folder)
+        set_folder(folder, change_extension(input_filename, extension))
+      end
     end
   end
 end

data/lib/bioinform/data_models/collection.rb

@@ -2,7 +2,10 @@ require 'ostruct'
 
 module Bioinform
   class Collection
-    attr_reader :collection, :parameters
+    attr_reader :collection
+
+    include Parameters
+    make_parameters :name
 
     # collection name is a tag name for each motif in a collection. But motif can be included in several collections so have several tags
     def initialize(parameters = {})
@@ -15,14 +18,14 @@ module Bioinform
       collection.size
     end
 
-    def name
-      parameters.name
+    def to_s(with_name = true)
+      result = (with_name) ? "Collection: #{name.to_s}\n" : ''
+      each do |pm, infos|
+        result << pm.to_s << "\n\n"
+      end
+      result
     end
 
-    def to_s
-      "<Collection '#{name}'>"
-    end
-    
     def +(other)
       result = self.class.new
       each do |pm, infos|
@@ -39,7 +42,7 @@ module Bioinform
       collection << [pm, info]
       self
     end
-    
+
     def <<(pm)
       add_pm(pm, OpenStruct.new)
     end
@@ -51,7 +54,7 @@ module Bioinform
         Enumerator.new(self, :each)
       end
     end
-    
+
     def each_pm
       if block_given?
         each{|pm, infos| yield pm}
@@ -75,7 +78,7 @@ module Bioinform
         end                                           #   end
       end                                             # end
     end
-    
+
     def ==(other)
       (collection == other.collection) && (parameters == other.parameters)
     rescue

data/lib/bioinform/data_models/pcm.rb

@@ -1,5 +1,5 @@
-require 'bioinform/support'
-require 'bioinform/data_models'
+require_relative '../support'
+require_relative '../data_models'
 
 module Bioinform
   class PCM < PM

data/lib/bioinform/data_models/pm.rb

@@ -1,13 +1,16 @@
-require 'bioinform/support'
-require 'bioinform/parsers'
+require 'ostruct'
+require_relative '../support'
+require_relative '../parsers'
 
 module Bioinform
   IndexByLetter = {'A' => 0, 'C' => 1, 'G' => 2, 'T' => 3, A: 0, C: 1, G: 2, T: 3}
   LetterByIndex = {0 => :A, 1 => :C, 2 => :G, 3 => :T}
 
   class PM
-    attr_reader :matrix, :tags
-    attr_accessor :background, :name
+    attr_accessor :matrix, :parameters
+
+    include Parameters
+    make_parameters :tags, :name, :background
 
     def mark(tag)
       tags << tag
@@ -18,24 +21,31 @@ module Bioinform
     end
 
     def self.choose_parser(input)
-      [TrivialParser, Parser, StringParser, StringFantomParser].find do |parser|
+      [TrivialParser, YAMLParser, Parser, StringParser, StringFantomParser, TrivialCollectionParser, YAMLCollectionParser].find do |parser|
         self.new(input, parser) rescue nil
       end
     end
+    
+    def self.split_on_motifs(input)
+      parser = choose_parser(input)
+      raise ParsingError, "No parser can parse given input"  unless parser
+      parser.split_on_motifs(input, self)
+    end
 
     def initialize(input, parser = nil)
+      @parameters = OpenStruct.new
       parser ||= self.class.choose_parser(input)
       raise 'No one parser can process input'  unless parser
       result = parser.new(input).parse
-      @matrix = result[:matrix]
-      @name = result[:name]
-      @tags = result[:tags] || []
-      @background = result[:background] || [1, 1, 1, 1]
+      @matrix = result.matrix
+      self.name = result.name
+      self.tags = result.tags || []
+      self.background = result.background || [1, 1, 1, 1]
       raise 'matrix not valid'  unless valid?
     end
 
     def ==(other)
-      @matrix == other.matrix && @background == other.background && @name == other.name
+      @matrix == other.matrix && background == other.background && name == other.name
     rescue
       false
     end
@@ -77,8 +87,8 @@ module Bioinform
         matrix_str = each_position.map{|pos| pos.join("\t")}.join("\n")
       end
 
-      if options[:with_name] && @name
-        @name + "\n" + matrix_str
+      if options[:with_name] && name
+        name + "\n" + matrix_str
       else
         matrix_str
       end
@@ -97,8 +107,8 @@ module Bioinform
 
       matrix_str = matrix_rows.join("\n")
 
-      if options[:with_name] && @name
-        @name + "\n" + header + matrix_str
+      if options[:with_name] && name
+        name + "\n" + header + matrix_str
       else
         header + matrix_str
       end
@@ -111,22 +121,6 @@ module Bioinform
       hsh.with_indifferent_access
     end
 
-    # pm.background - returns a @background attribute
-    # pm.background(new_background) - sets an attribute and returns pm itself
-    # if more than one argument passed - raises an exception
-    def background(*args)
-      case args.size
-      when 0 then @background
-      when 1 then background!(args[0])
-      else raise ArgumentError, '#background method can get 0 or 1 argument'
-      end
-    end
-
-    def background!(new_background)
-      @background = new_background
-      self
-    end
-
     def self.zero_column
       [0, 0, 0, 0]
     end
@@ -158,13 +152,6 @@ module Bioinform
       background.map{|element| element.to_f / sum}
     end
 
-    #def split(first_chunk_length)
-    #  [@matrix.first(first_chunk_length), matrix.last(length - first_chunk_length)]
-    #end
-    #def permute_columns(permutation_index)
-    #  @matrix.values_at(permutation_index)permutation_index.map{|col| matrix[col]}
-    #end
-
     def best_score
       @matrix.inject(0.0){|sum, col| sum + col.max}
     end

data/lib/bioinform/data_models/ppm.rb

@@ -1,5 +1,5 @@
-require 'bioinform/support'
-require 'bioinform/data_models'
+require_relative '../support'
+require_relative '../data_models'
 
 module Bioinform
   class PPM < PM

data/lib/bioinform/data_models/pwm.rb

@@ -1,5 +1,5 @@
-require 'bioinform/support'
-require 'bioinform/data_models'
+require_relative '../support'
+require_relative '../data_models'
 module Bioinform
   class PWM < PM
     def score_mean

data/lib/bioinform/data_models.rb

@@ -1,11 +1,11 @@
-require 'bioinform/parsers'
+require_relative 'parsers'
 
-require 'bioinform/data_models/pm'
-require 'bioinform/data_models/pcm'
-require 'bioinform/data_models/ppm'
-require 'bioinform/data_models/pwm'
+require_relative 'data_models/pm'
+require_relative 'data_models/pcm'
+require_relative 'data_models/ppm'
+require_relative 'data_models/pwm'
 
-require 'bioinform/data_models/collection'
+require_relative 'data_models/collection'
 
-#require 'bioinform/data_models/iupac_word'
-#require 'bioinform/data_models/iupac_wordset'
+#require_relative 'bioinform/data_models/iupac_word'
+#require_relative 'bioinform/data_models/iupac_wordset'

data/lib/bioinform/parsers/parser.rb

@@ -1,7 +1,13 @@
-require 'bioinform/support'
-require 'bioinform/data_models/pm'
+require 'ostruct'
+require_relative '../support'
+require_relative '../data_models/pm'
+require_relative 'splittable_parser'
 
 module Bioinform
+  class Error < StandardError; end
+  class ParsingError < Error; end
+  class InvalidMatrix < Error; end
+  
   class Parser
     attr_reader :input
 
@@ -19,8 +25,8 @@ module Bioinform
 
     def parse!
       matrix = self.class.transform_input(input)
-      raise 'Parsing error' unless self.class.valid_matrix?(matrix)
-      {matrix: matrix}
+      raise InvalidMatrix unless self.class.valid_matrix?(matrix)
+      OpenStruct.new(matrix: matrix)
     end
 
     def parse
@@ -78,6 +84,5 @@ module Bioinform
     def self.need_tranpose?(input)
       (input.size == 4) && input.any?{|x| x.size != 4}
     end
-
   end
 end

data/lib/bioinform/parsers/splittable_parser.rb

@@ -0,0 +1,57 @@
+module Bioinform
+  class Parser
+    module SingleMotifParser
+      def self.included(base)
+        base.class_eval { extend ClassMethods }
+        include Enumerable
+        alias_method :split, :to_a
+      end
+      module ClassMethods
+        def split_on_motifs(input, pm_klass = PM)
+          [ input.is_a?(pm_klass) ? self : pm_klass.new(input, self) ]
+        end
+      end
+      def each
+        if block_given?
+          yield self
+        else
+          Enumerator.new(self, :each)
+        end
+      end
+    end
+    include SingleMotifParser
+
+    module MultipleMotifsParser
+      def self.included(base)
+        base.class_eval { extend ClassMethods }
+        include Enumerable
+        alias_method :split, :to_a
+      end
+      module ClassMethods
+        def split_on_motifs(input, pm_klass = PM)
+          split(input).map{|el| el.is_a?(pm_klass) ? el : pm_klass.new(el)}
+        end
+        def split(input)
+          self.new(input).split
+        end
+        private :split
+      end
+
+      def scanner_reset
+      end
+
+      def each
+        if block_given?
+          scanner_reset
+          while result = parse
+            yield result
+          end
+        else
+          Enumerator.new(self, :each)
+        end
+      end
+
+      private :scanner_reset
+    end
+  end
+end

data/lib/bioinform/parsers/string_fantom_parser.rb

@@ -1,5 +1,5 @@
-require 'bioinform/support'
-require 'bioinform/parsers/string_parser'
+require_relative '../support'
+require_relative '../parsers/string_parser'
 
 module Bioinform
   class StringFantomParser < StringParser
@@ -28,7 +28,7 @@ module Bioinform
       scan_splitter
       name = parse_name
       matrix = parse_matrix
-      Parser.parse!(matrix).merge(name: name)
+      Parser.parse!(matrix).tap{|result| result.name = name}
     end
 
   end

data/lib/bioinform/parsers/string_parser.rb

@@ -1,10 +1,12 @@
 require 'strscan'
-require 'bioinform/support'
-require 'bioinform/parsers/parser'
+require_relative '../support'
+require_relative '../parsers/parser'
 
 module Bioinform
   class StringParser < Parser
+    include MultipleMotifsParser
     attr_reader :scanner, :row_acgt_markers
+
     def initialize(input)
       raise ArgumentError  unless input.is_a?(String)
       super
@@ -60,32 +62,11 @@ module Bioinform
       parse_acgt_header
       matrix = parse_matrix
       matrix = matrix.transpose if row_acgt_markers
-      Parser.parse!(matrix).merge(name: name)
+      Parser.parse!(matrix).tap{|result| result.name = name}
     end
 
     def scanner_reset
       scanner.reset
     end
-
-    def each
-      if block_given?
-        scanner_reset
-        while result = parse
-          yield result
-        end
-      else
-        Enumerator.new(self, :each)
-      end
-    end
-    include Enumerable
-
-    alias_method :split, :to_a
-    def self.split(input)
-      self.new(input).split
-    end
-
-    def self.split_on_motifs(input, pm_klass = PM)
-      split(input).map{|el| pm_klass.new(el)}
-    end
   end
 end