obo_parser 0.3.6 → 0.3.7

data/VERSION

@@ -1 +1 @@
-0.3.6
+0.3.7

data/lib/utilities.rb

@@ -18,7 +18,7 @@ module OboParser::Utilities
   #
   # @param [Integer] cutoff only Term ids with > cutoff labels will be reported 
   # @param [Array] files an Array of read files 
-  # @return [String] the transation in tab delimted format
+  # @return [String] the translation in tab delimted format
   def self.dump_comparison_by_id(cutoff = 0, files = [])
     return '' if files.size < 1
 
@@ -47,7 +47,9 @@ module OboParser::Utilities
     end
   end
 
+
   # Returns all labels found in all passed ontologies. Does not yet include synonyms.
+  # Caution: strips adult, embryonic, larval from labels (comment to remove)
   #
   #== Example use
   #  of1 = File.read('fly_anatomy.obo')	
@@ -64,8 +66,11 @@ module OboParser::Utilities
     files.each do |f|
       o = parse_obo_file(f)
       o.term_hash.keys.each do |k|
+
+        # TODO: make this optional 
         tmp = k.gsub(/adult/, "").strip
         tmp = k.gsub(/embryonic\/larval/, "").strip
+
         if comparison[tmp]
           comparison[tmp] += 1
         else
@@ -86,10 +91,9 @@ module OboParser::Utilities
 
   end 
 
+  #== Two column correspondances and translation tools
 
-  #== Two column translation tools
-
-HOMOLONTO_HEADER = %{
+  HOMOLONTO_HEADER = %{
 format-version: 1.2
 auto-generated-by: obo_parser
 default-namespace: fix_me
@@ -212,31 +216,129 @@ is_anti_symmetric: true
     true
   end
 
-  # Takes a two column input file, references it to two ontologies, and returns a hash
+
+  # Takes a two column input file, references it to two ontologies, and returns a report 
+  # that identifies data pairs that have parents who are also a data pair given a 
+  # provided property/relation type.
   #  
   #== Example use
   #  file = File.read('HAO_TGMA_list.txt')
   #  col1_obo = File.read('hao.obo')
   #  col2_obo = File.read('tgma.obo')
+  #
+  # foo = OboParser::Utilities.parents(:data => data, :col1_obo => col1_obo, :col2_obo => col2_obo, :property => 'is_a')
+  #
+  # puts "-- NO (#{foo[:no].size})\n" 
+  # puts foo[:no].join("\n")
+  # puts "-- YES (#{foo[:yes].size})\n" 
+  # puts foo[:yes].join("\n")
+  #
+  # @param [Hash] options options.
+  # @param [Symbol] data the two column data file.
+  # @param [Symbol] colo1_obo the OBO file referenced in the first column 
+  # @param [Symbol] colo2_obo the OBO file referenced in the second column 
+  # @param [Symbol] property the OBO relationship/property to check against (e.g. 'is_a', 'part_of') 
+  # @return [Hash] a hash of {:yes => {}, :no => {}}
+  def parents(options = {})
+    opt = {
+      :data => nil,
+      :col1_obo => nil,
+      :col2_obo => nil,
+      :property => nil
+    }.merge!(options)
+
+    return false if opt[:property].nil? 
+    c1obo = parse_obo_file(opt[:col1_obo])
+    c2obo = parse_obo_file(opt[:col2_obo])
+
+    result = {:yes => [], :no => [], :unplaced => []}
+    # update
+    array = OboParser::Utilities.arrayify_pairs(:data => opt[:data], :col1_obo => opt[:col1_obo], :col2_obo => opt[:col2_obo])
+
+    # index the pairs
+    index = array.inject({}){|hsh, i| hsh.update({"#{i[0]}#{i[1]}" => i})}
+
+    obo1_hash = c1obo.id_index
+    obo2_hash = c2obo.id_index
+
+    array.each do |k|
+      a = k[0]
+      b = k[1]
+
+      ids_1 = []
+      ids_2 = []
+
+      if !obo1_hash[a]
+        puts "can't find #{k}\n"
+        next
+      end
+
+      if !obo2_hash[b]
+        puts "can't find #{k}\n"
+        next
+      end
+
+      obo1_hash[a].relationships.each do |rel, id| 
+        if rel == opt[:property] 
+          ids_1.push id
+        end
+      end
+
+      obo2_hash[b].relationships.each do |rel, id|
+        if rel == opt[:property] 
+          ids_2.push id
+        end
+      end
+
+      unplaced = true
+
+      ids_1.each do |c|
+        ids_2.each do |d|
+          t = "#{a} -> #{b}"
+          if index["#{c}#{d}"]   # think this is right   hash[c] == d
+            result[:yes].push(t)
+            unplaced = false
+            next # don't add again after we find a hit
+          else
+            result[:no].push(t)
+            unplaced = false
+          end
+        end
+      end
+      result[:unplaced] 
+
+    end
+
+    result
+  end
+
+
+  # Takes a two column input file, references it to two ontologies, and returns a array of [[id1, id2], [id3,idn] ...].
+  #
+  #== Example use
+  #  file = File.read('HAO_TGMA_list.txt')
+  #  col1_obo = File.read('hao.obo')
+  #  col2_obo = File.read('tgma.obo')
   #  
-  #  OboParser::Utilities.hashify_pairs(:data => file, :col1_obo => col1_obo, :col2_obo => col2_obo)
+  #  OboParser::Utilities.arrayify_pairs(:data => file, :col1_obo => col1_obo, :col2_obo => col2_obo)
   #
   # @param [Hash] options options.
   # @param [Symbol] data the two column data file.
   # @param [Symbol] colo1_obo the OBO file referenced in the first column 
   # @param [Symbol] colo2_obo the OBO file referenced in the second column 
-  # @return [Hash] a hash of {id string => id string}
-  def self.hashify_pairs(options = {})
+  # @return [Array] an array of [[id string, id string], ...]  
+  def self.arrayify_pairs(options = {})
     opt = {
       :data => nil,
       :col1_obo => nil,
       :col2_obo => nil,
+      :index_start => 0
     }.merge!(options)
-    
+
     c1obo = parse_obo_file(opt[:col1_obo])
     c2obo = parse_obo_file(opt[:col2_obo])
 
-    hash = Hash.new
+    array = Array.new
 
     i = opt[:index_start]
     v1 = nil # a label like 'head'
@@ -246,7 +348,7 @@ is_anti_symmetric: true
 
     opt[:data].split(/\n/).each do |row|
       i += 1
-      c1, c2 =  row.split(/\t/).map(&:strip)
+      c1, c2 = row.split(/\t/).map(&:strip)
 
       if c1.nil? || c2.nil?
         next
@@ -263,11 +365,78 @@ is_anti_symmetric: true
       else
         v2 = c2obo.term_hash[c2]
       end
-   
-     hash.merge!(c1 => c2) 
-    
+
+      array << [c1,c2] 
+
+    end
+    return array 
+  end
+
+
+  # Provides a set comparisson of correspondances between two sets of correspondances.  Finds correspondances that are shared, in left only, or in right only.
+  #  
+  #== Example use
+  #  data1 = File.read('input/hao_fbbt_vetted_list2.txt')     
+  #  data2 = File.read('input/hao_fbbt_from_bioportal.txt')   
+  #  col1_obo = File.read('hao.obo')
+  #  col2_obo = File.read('tgma.obo')
+  #  
+  #  OboParser::Utilities.set_comparison(:data1 => data1 , :data2 => data2, :col1_obo => col1_obo, :col2_obo => col2_obo, :return => :text ) # => a text report 
+  # or
+  #  foo = OboParser::Utilities.set_comparison(:data1 => data1 , :data2 => data2, :col1_obo => col1_obo, :col2_obo => col2_obo, :return => :array ) 
+  #
+  #== "Report" types 
+  # Use :return to return different values 
+  #   :array - Returns an Array
+  #   :text - Generates 4 column report (id, label, id, label) to STDOUT
+  #
+  # @param [Hash] options options.
+  # @param [Symbol] data1 two columns (tab) with a correspondance between ontologies 1 and 2
+  # @param [Symbol] data2 as data1, a second set of correspondances 
+  # @param [Symbol] col1_obo the OBO file corresponding to the first column of data1 and data2
+  # @param [Symbol] col2_obo the OBO file corresponding to the second column of data1 and data2
+  # @param [Symbol] return the value to return
+  # @return [String | Array]  
+  def self.set_comparison(options = {})
+    opt = {
+      :data1 => nil,
+      :data2 => nil,
+      :col1_obo => nil,
+      :col2_obo => nil,
+      :return => :array # also :text
+    }.merge!(options)
+
+    c1obo = parse_obo_file(opt[:col1_obo])
+    c2obo = parse_obo_file(opt[:col2_obo])
+
+    th1 = c1obo.id_hash
+    th2 = c2obo.id_hash
+
+    array1 =  OboParser::Utilities.arrayify_pairs(:data => opt[:data1], :col1_obo => opt[:col1_obo], :col2_obo => opt[:col2_obo])
+    array2 =  OboParser::Utilities.arrayify_pairs(:data => opt[:data2], :col1_obo => opt[:col1_obo], :col2_obo => opt[:col2_obo])
+
+    index1 = array1.inject({}){|hsh, i| hsh.merge!( "#{i[0]}#{i[1]}" => i)}
+    index2 = array2.inject({}){|hsh, i| hsh.merge!( "#{i[0]}#{i[1]}" => i)}
+
+    result = {:in_left => [], :in_right => [], :in_both => []}
+
+    result[:in_both]  = (index1.keys & index2.keys).collect{|k| [ index1[k][0], th1[index1[k][0]], index1[k][1], th2[index1[k][1]]                   ,             ]  }
+    result[:in_left]  = ((index1.keys | index2.keys)  - index2.keys ).collect{|k| [ index1[k][0], th1[index1[k][0]], index1[k][1], th2[index1[k][1]] ,             ]  }
+    result[:in_right] = ((index1.keys | index2.keys)  - index1.keys ).collect{|k| [ index2[k][0], th1[index2[k][0]], index2[k][1], th2[index2[k][1]] ,             ]  }
+
+    case opt[:return]
+    when :array
+      return result
+    when :text
+      [:in_both, :in_left, :in_right].each do |t|
+        puts "--- #{t.to_s} (#{result[t].size}) ---"
+        result[t].each do |r|
+          puts r.join("\t")
+        end
+      end
+    else
+      return nil
     end
-    return hash
   end
 
 
@@ -290,143 +459,48 @@ is_anti_symmetric: true
   end
 
 
-# Takes a Hash of OBO ontology files, an Array of relationships, and writes two input files (a network, and node properties) for Cytoscape
-#  
-#== Example use
-# OboParser::Utilities.cytoscapify(:ontologies => {'HAO' => File.read('input/hao.obo'), 'TADS' => File.read('input/tads.obo'), 'TGMA' => File.read('input/tgma.obo'), 'FBBT' => File.read('input/fbbt.obo') }, :properties => ['is_a', 'part_of'])
-#
-# @param [Symbol] ontologies a Hash of #read files as values, keys as working names
-# @param [Symbol] properties an Array of properties like ['is_a', 'part_of'] 
-# TODO: @return File1, File2, Filen  
-def self.cytoscapify(options = {})
-  opt = {
-    :ontologies => {},
-    :properties => []
-  }.merge!(options)
-
-  return false if opt[:properties].empty?
-  return false if opt[:ontologies].empty?
-
-  nodes = File.new("nodes.tab", "w+")
-  edges = File.new("edges.eda", "w+")
-
-  opt[:ontologies].keys.each do |k|
-
-    obo_file = parse_obo_file(opt[:ontologies][k])
-
-    obo_file.terms.each do |t|
-      nodes.puts [t.id.value, t.name.value, k].join("\t") + "\n"
-
-      t.relationships.each do |rel, id|
-        edges.puts [t.id.value, "(#{rel})", id].join("\t") + "\n" if opt[:properties].include?(rel)
-      end
-    end
-  end
-
-  nodes.close
-  edges.close
-
-  true
-
-end
-
+  # Takes a Hash of OBO ontology files, an Array of relationships, and writes two input files (a network, and node properties) for Cytoscape
+  #  
+  #== Example use
+  # OboParser::Utilities.cytoscapify(:ontologies => {'HAO' => File.read('input/hao.obo'), 'TADS' => File.read('input/tads.obo'), 'TGMA' => File.read('input/tgma.obo'), 'FBBT' => File.read('input/fbbt.obo') }, :properties => ['is_a', 'part_of'])
+  #
+  # @param [Symbol] ontologies a Hash of #read files as values, keys as working names
+  # @param [Symbol] properties an Array of properties like ['is_a', 'part_of'] 
+  # TODO: @return File1, File2, Filen  
+  def self.cytoscapify(options = {})
+    opt = {
+      :ontologies => {},
+      :properties => []
+    }.merge!(options)
 
-# Takes a two column input file, references it to two ontologies, and returns a report 
-# that identifies data pairs that have parents who are also a data pair given a 
-# provided property/relation type.
-#  
-#== Example use
-#  file = File.read('HAO_TGMA_list.txt')
-#  col1_obo = File.read('hao.obo')
-#  col2_obo = File.read('tgma.obo')
-#
-# foo = OboParser::Utilities.parents(:data => data, :col1_obo => col1_obo, :col2_obo => col2_obo, :property => 'is_a')
-#
-# puts "-- NO (#{foo[:no].size})\n" 
-# puts foo[:no].join("\n")
-# puts "-- YES (#{foo[:yes].size})\n" 
-# puts foo[:yes].join("\n")
-#
-# @param [Hash] options options.
-# @param [Symbol] data the two column data file.
-# @param [Symbol] colo1_obo the OBO file referenced in the first column 
-# @param [Symbol] colo2_obo the OBO file referenced in the second column 
-# @param [Symbol] property the OBO relationship/property to check against (e.g. 'is_a', 'part_of') 
-# @return [Hash] a hash of {:yes => {}, :no => {}}
-def self.parents(options = {})
-  opt = {
-    :data => nil,
-    :col1_obo => nil,
-    :col2_obo => nil,
-    :property => nil
-  }.merge!(options)
-
-  return false if opt[:property].nil? 
-  c1obo = parse_obo_file(opt[:col1_obo])
-  c2obo = parse_obo_file(opt[:col2_obo])
-
-  result = {:yes => [], :no => [], :unplaced => []}
-  # update
-  hash =  hashify_pairs(:data => opt[:data], :col1_obo => opt[:col1_obo], :col2_obo =>  opt[:col2_obo])
-
-  obo1_hash = c1obo.id_index
-  obo2_hash = c2obo.id_index
-
-  hash.keys.each do |k|
-    a = k
-    b = hash[a]
-
-    ids_1 = []
-    ids_2 = []
-
-    if !obo1_hash[a]
-      puts "can't find #{k}\n"
-      next
-    end
+    return false if opt[:properties].empty?
+    return false if opt[:ontologies].empty?
 
-    if !obo2_hash[b]
-      puts "can't find #{k}\n"
-      next
-    end
+    nodes = File.new("nodes.tab", "w+")
+    edges = File.new("edges.eda", "w+")
 
-    obo1_hash[a].relationships.each do |rel, id| 
-      if rel == opt[:property] 
-        ids_1.push id
-      end
-    end
+    opt[:ontologies].keys.each do |k|
 
-    obo2_hash[b].relationships.each do |rel, id|
-      if rel == opt[:property] 
-        ids_2.push id
-      end
-    end
+      obo_file = parse_obo_file(opt[:ontologies][k])
 
-    unplaced = true
+      obo_file.terms.each do |t|
+        nodes.puts [t.id.value, t.name.value, k].join("\t") + "\n"
 
-    ids_1.each do |c|
-      ids_2.each do |d|
-        t = "#{a} -> #{b}"
-        if hash[c] == d
-          result[:yes].push(t)
-          unplaced = false
-          next # don't add again after we find a hit
-        else
-          result[:no].push(t)
-          unplaced = false
+        t.relationships.each do |rel, id|
+          edges.puts [t.id.value, "(#{rel})", id].join("\t") + "\n" if opt[:properties].include?(rel)
         end
       end
     end
-    result[:unplaced] 
 
-  end
-
-  result
-end
+    nodes.close
+    edges.close
 
+    true
 
+  end
 
 
-#== Helper methods that don't require the obo_parser library
+  #== Helper methods that don't require the obo_parser library
 
   # Given a Term id and a String representing an OBO file returns that stanza. 
   #

data/obo_parser.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = %q{obo_parser}
-  s.version = "0.3.6"
+  s.version = "0.3.7"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["mjy"]
-  s.date = %q{2011-08-10}
+  s.date = %q{2011-08-25}
   s.description = %q{Provides all-in-one object containing the contents of an OBO formatted file.  OBO version 1.2 is targeted, though this should work for 1.0. }
   s.email = %q{diapriid@gmail.com}
   s.extra_rdoc_files = [

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: obo_parser
 version: !ruby/object:Gem::Version 
-  hash: 31
+  hash: 29
   prerelease: 
   segments: 
   - 0
   - 3
-  - 6
-  version: 0.3.6
+  - 7
+  version: 0.3.7
 platform: ruby
 authors: 
 - mjy
@@ -15,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-08-10 00:00:00 Z
+date: 2011-08-25 00:00:00 Z
 dependencies: []
 
 description: "Provides all-in-one object containing the contents of an OBO formatted file.  OBO version 1.2 is targeted, though this should work for 1.0. "