rbbt-GE 0.2.0 → 1.0.0

data/lib/rbbt/GE/GEO.rb

@@ -1,3 +1,4 @@
+# encoding: utf-8
 require 'rbbt-util'
 require 'rbbt/GE'
 require 'rbbt/sources/organism'
@@ -9,7 +10,7 @@ module GEO
   self.pkgdir = "geo"
   self.subdir = "arrays"
 
-  GEO.claim GEO.root.find(:user), :rake, Rbbt.share.install.GEO.Rakefile.find(:lib)
+  GEO.claim GEO.root, :rake, Rbbt.share.install.GEO.Rakefile.find(:lib)
 
   def self.comparison_name(field, condition, control)
     condition = condition * " AND " if Array === condition
@@ -74,11 +75,12 @@ module GEO
     }
    
     GDS_INFO = {
-      :DELIMITER        => "\\^SUBSET|!sample_table_begin",
-      :title         => "!Sample_title",
-      :accession         => "!Sample_geo_accession",
-      :channel_count    => "!Sample_channel_count",
-      :platform         => "!Sample_platform_id",
+      :DELIMITER        => "\\^SUBSET",
+      :value_type       => "!dataset_value_type",
+      :channel_count    => "!dataset_channel_count",
+      :platform         => "!dataset_platform",
+      :reference_series => "!dataset_reference_series",
+      :description      => "!dataset_description"
     }
 
     GDS_SUBSET_INFO = {
@@ -149,7 +151,7 @@ module GEO
     end
 
     def self.guess_id(organism, codes)
-      num_codes = codes.size
+      num_codes = codes.length
       best = nil
       best_count = 0
       new_fields = []
@@ -160,9 +162,15 @@ module GEO
           values = CMD.cmd("cat #{ codefile }|cut -f #{ i + 1 }| tr '|' '\\n'|grep [[:alpha:]]|sort -u").read.split("\n").reject{|code| code.empty?}
 
           new_field, count =  Organism.guess_id(organism, values)
+          new_field ||= field
+          count ||= 0
+          new_field = "UNKNOWN(#{new_field})" unless count > (num_codes > 20000 ? 20000 : num_codes).to_f * 0.2 and count > values.uniq.length * 0.5
+
+          Log.debug "Original field: #{ field }. New: #{new_field}. Count: #{ count }/#{num_codes}/#{values.uniq.length}"
+          new_fields << new_field
+
           field_counts[new_field] = count
-          Log.debug "Original field: #{ field }. New: #{new_field}. Count: #{ count }/#{num_codes}"
-          new_fields << (count > (num_codes > 20000 ? 20000 : num_codes).to_f * 0.5 ? new_field : "UNKNOWN(#{ field })")
+
           if count > best_count
             best = new_field
             best_count = count
@@ -184,7 +192,8 @@ module GEO
       code_file = File.join(directory, 'codes') 
       info_file = File.join(directory, 'info.yaml') 
 
-      stream = Open.open(GPL_URL.gsub('#PLATFORM#', platform), :nocache => true, :pipe => true)
+      # Fix platforms with the '.\d' extension (eg. NM_020527.1)
+      stream = CMD.cmd('sed \'s/\.[[:digit:]]\+\(\t\|$\)/\1/g;s/ *\/\/[^\t]*//g\'', :in =>  Open.open(GPL_URL.gsub('#PLATFORM#', platform), :nocache => true), :pipe => true)
 
       info = parse_header(stream, GPL_INFO)
       info[:code_file]      = code_file
@@ -278,7 +287,9 @@ module GEO
 
         samples << sample
 
-        sample_values = TSV.open(StringIO.new(chunk.match(/!sample_table_begin(.*)!sample_table_end/msi)[0].strip), :type => :list, :header_hash => '')
+        chunk = chunk.encode "UTF-8"
+        chunk = Misc.fixutf8 chunk
+        sample_values = TSV.open(StringIO.new(chunk.match(/!sample_table_begin\n(.*)\n!sample_table_end/mi)[1].strip), :type => :list, :header_hash => '')
         sample_values.fields = [sample]
 
         if values.nil?
@@ -310,12 +321,14 @@ module GEO
       key_field = TSV.parse_header(GEO[info[:platform]]['codes'].open).key_field
       values.key_field = key_field
 
-      Open.write(value_file, values.to_s)
-      Open.write(info_file, info.to_yaml)
-
+      info[:sample_info] ||= sample_info
       info[:channel_count] ||= sample_info.values.first[:channel_count]
       info[:value_type] ||= sample_info.values.first[:value_type]
 
+
+      Open.write(value_file, values.to_s)
+      Open.write(info_file, info.to_yaml)
+
       info
     end
   end
@@ -346,3 +359,4 @@ module GEO
     GE.analyze(value_file, condition_samples, control_samples, log2, path, format)
   end
 end
+

data/lib/rbbt/GE.rb

@@ -9,15 +9,18 @@ module GE
     R.run(cmd)
   end
 
-  def self.r_format(list)
+  def self.r_format(list, options = {})
+    strings = options[:strings]
     case
     when list.nil?
       "NULL"
     when Array === list
-      "c(#{list.collect{|e| r_format e} * ", "})"
-    when (String === list and list === list.to_i.to_s) 
+      "c(#{list.collect{|e| r_format e, options} * ", "})"
+    when (Fixnum === list or Float === list)
+      list.to_s
+    when (not strings and String === list and list === list.to_i.to_s) 
       list.to_i
-    when (String === list and list === list.to_f.to_s) 
+    when (not strings and String === list and list === list.to_f.to_s) 
       list.to_f
     when TrueClass === list
       "TRUE"
@@ -28,9 +31,9 @@ module GE
     end
   end
 
-  def self.analyze(datafile,  main, contrast = nil, log2 = false, outfile = nil, key_field = nil)
+  def self.analyze(datafile,  main, contrast = nil, log2 = false, outfile = nil, key_field = nil, two_channel = nil)
     FileUtils.mkdir_p File.dirname(outfile) unless outfile.nil? or File.exists? File.dirname(outfile)
-    GE.run_R("rbbt.GE.process(#{ r_format datafile }, main = #{r_format(main)}, contrast = #{r_format(contrast)}, log2=#{ r_format log2 }, outfile = #{r_format outfile}, key.field = #{r_format key_field})")
+    GE.run_R("rbbt.GE.process(#{ r_format datafile }, main = #{r_format(main, :strings => true)}, contrast = #{r_format(contrast, :strings => true)}, log2=#{ r_format log2 }, outfile = #{r_format outfile}, key.field = #{r_format key_field}, two.channel = #{r_format two_channel})")
   end
 end
 

data/share/install/GEO/Rakefile

@@ -6,6 +6,26 @@ rule /^(GPL\d+)\/?(codes|info\.yaml)?$/ do |t|
   GEO::SOFT.GPL(platform, file.nil? ? t.name : File.dirname(t.name))
 end
 
+rule /^(GDS\d+)\/samples$/ => [proc{|t| t.sub('samples', 'info.yaml')}, proc{|t| t.sub('samples', 'values')} ] do |t|
+  info = YAML.load(Open.open(t.prerequisites.first))
+
+  subsets = info[:subsets]
+  fields = subsets.keys
+
+  all_samples = TSV::Parser.new(Open.open(t.prerequisites.last)).fields
+  samples = TSV.setup(all_samples, :key_field => "Sample", :fields => [], :type => :list)
+
+  fields.each do |field|
+    assignments = subsets[field]
+
+    samples = samples.add_field field do |sample, v|
+      assignments.select{|value, list| list.include? sample}.first.first
+    end
+  end
+  
+  Open.write(t.name, samples.to_s)
+end
+
 rule /^(GDS\d+)\/?(values|info\.yaml)?$/ do |t|
   t.name =~ /^(GDS\d+)\/?(values|info\.yaml)?/
   dataset = $1

data/share/lib/R/MA.R

@@ -55,13 +55,31 @@ rbbt.GE.process.limma.twoside <- function(expr, subset.main, subset.contrast){
 }
 
 
+rbbt.GE.guess.log2 <- function(m, two.channel){
+    if (two.channel){
+        return (sum(m < 0, na.rm = TRUE) == 0);
+    }else{
+        return (max(m, na.rm = TRUE) > 100);
+    }
+}
+
+
 
-rbbt.GE.process <- function(file, main, contrast = NULL, log2 = FALSE, outfile = NULL, key.field = NULL){
+rbbt.GE.process <- function(file, main, contrast = NULL, log2 = FALSE, outfile = NULL, key.field = NULL, two.channel = NULL){
     data = rbbt.tsv(file);
     ids = rownames(data);
 
+    if (is.null(log2)){
+      print(str(data));
+      log2 = rbbt.GE.guess.log2(data, two.channel)
+    }
+
     if (log2){
        data = log2(data);
+       min = min(data[data != -Inf])
+       data[data == -Inf] = min
+       print(summary(data));
+       return
     }
 
     if (is.null(contrast)){
@@ -103,6 +121,7 @@ rbbt.GE.process <- function(file, main, contrast = NULL, log2 = FALSE, outfile =
 
     if (! is.null(limma)){
        result = data.frame(ratio = ratio[ids], t.values = limma$t[ids], p.values = limma$p.values[ids])
+       result["adjusted.p.values"] = p.adjust(result$p.values, "fdr")
     }else{
        result = data.frame(ratio = ratio)
     }

metadata

@@ -1,45 +1,39 @@
---- !ruby/object:Gem::Specification 
+--- !ruby/object:Gem::Specification
 name: rbbt-GE
-version: !ruby/object:Gem::Version 
-  hash: 23
+version: !ruby/object:Gem::Version
+  version: 1.0.0
   prerelease: 
-  segments: 
-  - 0
-  - 2
-  - 0
-  version: 0.2.0
 platform: ruby
-authors: 
+authors:
 - Miguel Vazquez
 autorequire: 
 bindir: bin
 cert_chain: []
-
-date: 2012-01-31 00:00:00 Z
-dependencies: 
-- !ruby/object:Gem::Dependency 
+date: 2012-12-21 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
   name: rbbt-util
-  prerelease: false
-  requirement: &id001 !ruby/object:Gem::Requirement 
+  requirement: !ruby/object:Gem::Requirement
     none: false
-    requirements: 
-    - - ">="
-      - !ruby/object:Gem::Version 
-        hash: 3
-        segments: 
-        - 0
-        version: "0"
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
   type: :runtime
-  version_requirements: *id001
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
 description: Gene Expression in RBBT
 email: miguel.vazquez@cnio.es
 executables: []
-
 extensions: []
-
-extra_rdoc_files: 
+extra_rdoc_files:
 - LICENSE
-files: 
+files:
 - LICENSE
 - lib/rbbt/GE.rb
 - lib/rbbt/GE/GEO.rb
@@ -49,37 +43,28 @@ files:
 - test/rbbt/GE/test_GEO.rb
 homepage: http://github.com/mikisvaz/rbbt-GE
 licenses: []
-
 post_install_message: 
 rdoc_options: []
-
-require_paths: 
+require_paths:
 - lib
-required_ruby_version: !ruby/object:Gem::Requirement 
+required_ruby_version: !ruby/object:Gem::Requirement
   none: false
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      hash: 3
-      segments: 
-      - 0
-      version: "0"
-required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      hash: 3
-      segments: 
-      - 0
-      version: "0"
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
 requirements: []
-
 rubyforge_project: 
-rubygems_version: 1.8.10
+rubygems_version: 1.8.24
 signing_key: 
 specification_version: 3
 summary: Gene Expression in RBBT
-test_files: 
+test_files:
 - test/test_helper.rb
 - test/rbbt/GE/test_GEO.rb