bio 1.4.1 → 1.4.2

data/lib/bio.rb

@@ -82,6 +82,7 @@ module Bio
   ## below are described in bio/db/genbank/ddbj.rb
   #class DDBJ
   #  autoload :XML,          'bio/io/ddbjxml'
+  #  autoload :REST,         'bio/io/ddbjrest'
   #end
 
   ## EMBL/TrEMBL/Swiss-Prot/SPTR

data/lib/bio/appl/blast.rb

@@ -30,7 +30,7 @@ module Bio
   #   
   #   # To run an actual BLAST analysis:
   #   #   1. create a BLAST factory
-  #   remote_blast_factory = Bio::Blast.remote('blastp', 'SWISS',
+  #   remote_blast_factory = Bio::Blast.remote('blastp', 'swissprot',
   #                                            '-e 0.0001', 'genomenet')
   #   #or:
   #   local_blast_factory = Bio::Blast.local('blastn','/path/to/db')

data/lib/bio/appl/blast/ddbj.rb

@@ -8,7 +8,7 @@
 #
 
 require 'bio/appl/blast/remote'
-require 'bio/io/ddbjxml'
+require 'bio/io/ddbjrest'
 
 module Bio::Blast::Remote
 
@@ -37,32 +37,43 @@ module Bio::Blast::Remote
         if defined? @parse_databases
           return nil if @parse_databases
         end
-        drv = Bio::DDBJ::XML::Blast.new
+        drv = Bio::DDBJ::REST::Blast.new
         str = drv.getSupportDatabaseList
 
         databases = {}
         dbdescs = {}
-        key = 'blastn'
+        keys = [ 'blastn', 'blastp' ]
+        keys.each do |key|
+          databases[key] ||= []
+          dbdescs[key] ||= {}
+        end
         prefix = ''
-        databases[key] ||= []
-        dbdescs[key] ||= {}
+        prefix_count = 0
         str.each_line do |line|
           a = line.strip.split(/\s*\-\s*/, 2)
           case a.size
           when 1
             prefix = a[0].to_s.strip
             prefix += ': ' unless prefix.empty?
-            key = 'blastn'
+            prefix_count = 0
             next #each_line
           when 0
-            prefix = ''
-            key = 'blastp'
-            databases[key] ||= []
-            dbdescs[key] ||= {}
+            prefix = '' if prefix_count > 0
             next #each_line
           end
           name = a[0].to_s.strip.freeze
-          desc = (prefix + a[1].to_s.strip).freeze
+          desc = a[1].to_s.strip
+          key = case desc
+                when /\(NT\)\s*$/
+                  'blastn'
+                when /\(AA\)\s*$/
+                  'blastp'
+                else
+                  warn "DDBJ BLAST: could not determine the database is NT or AA: #{line.chomp}" if $VERBOSE
+                  next #each_line
+                end
+          desc = (prefix + desc).freeze
+          prefix_count += 1
           databases[key].push name
           dbdescs[key][name] = desc
         end
@@ -96,11 +107,9 @@ module Bio::Blast::Remote
       options = make_command_line_options
       opt = Bio::Blast::NCBIOptions.new(options)
 
-      # SOAP objects are cached
-      @ddbj_remote_blast ||= Bio::DDBJ::XML::Blast.new
-      #@ddbj_request_manager ||= Bio::DDBJ::XML::RequestManager.new
-      # always use REST version to prevent warning messages
-      @ddbj_request_manager ||= Bio::DDBJ::XML::RequestManager::REST.new
+      # REST objects are cached
+      @ddbj_remote_blast ||= Bio::DDBJ::REST::Blast.new
+      @ddbj_request_manager ||= Bio::DDBJ::REST::RequestManager.new
 
       program = opt.delete('-p')
       db = opt.delete('-d')
@@ -110,24 +119,7 @@ module Bio::Blast::Remote
       qid = @ddbj_remote_blast.searchParamAsync(program, db, query, optstr)
       @output = qid
 
-      sleeptime = 2
-      flag = true
-      while flag
-        if $VERBOSE then
-          $stderr.puts "DDBJ BLAST: ID: #{qid} -- waitng #{sleeptime} sec."
-        end
-        sleep(sleeptime)
-
-        result = @ddbj_request_manager.getAsyncResult(qid)
-        case result.to_s
-        when /The search and analysis service by WWW is very busy now/
-          raise result.to_s.strip + '(Alternatively, wrong options may be given.)'
-        when /Your job has not completed yet/
-          sleeptime = 5
-        else
-          flag = false
-        end
-      end while flag
+      result = @ddbj_request_manager.wait_getAsyncResult(qid)
 
       @output = result
       return @output

data/lib/bio/appl/blast/genomenet.rb

@@ -69,11 +69,11 @@ module Bio::Blast::Remote
   # 
   # * http://www.ncbi.nlm.nih.gov/blast/
   # * http://www.ncbi.nlm.nih.gov/Education/BLASTinfo/similarity.html
-  # * http://blast.genome.jp/ideas/ideas.html#blast
+  # * http://www.genome.jp/tools/blast/
   #
   module GenomeNet
 
-    Host = "blast.genome.jp".freeze
+    Host = "www.genome.jp".freeze
 
     # Creates a remote BLAST factory using GenomeNet.
     # Returns Bio::Blast object.
@@ -100,7 +100,7 @@ module Bio::Blast::Remote
         key = nil
         host = Bio::Blast::Remote::Genomenet::Host
         http = Bio::Command.new_http(host)
-        result = http.get('/')
+        result = http.get('/tools/blast/')
         #p result.body
         result.body.each_line do |line|
           case line
@@ -166,11 +166,20 @@ module Bio::Blast::Remote
       program = opt.delete('-p')
       db = opt.delete('-d')
 
+      # When database name starts with mine-aa or mine-nt,
+      # space-separated list of KEGG organism codes can be given.
+      # For example, "mine-aa eco bsu hsa".
+      if /\A(mine-(aa|nt))\s+/ =~ db.to_s then
+        db = $1
+        myspecies = {}
+        myspecies["myspecies-#{$2}"] = $'
+      end
+
       matrix = opt.delete('-M') || 'blosum62'
       filter = opt.delete('-F') || 'T'
 
-      opt_V = opt.delete('-V') || 500 # default value for GenomeNet
-      opt_B = opt.delete('-B') || 250 # default value for GenomeNet
+      opt_v = opt.delete('-v') || 500 # default value for GenomeNet
+      opt_b = opt.delete('-b') || 250 # default value for GenomeNet
 
       # format, not for form parameters, but included in option string
       opt_m = opt.get('-m') || '7' # default of BioRuby GenomeNet factory
@@ -186,11 +195,13 @@ module Bio::Blast::Remote
         'other_param'    => optstr,
         'matrix'         => matrix,
         'filter'         => filter,
-        'V_value'        => opt_V, 
-        'B_value'        => opt_B, 
+        'V_value'        => opt_v, 
+        'B_value'        => opt_b, 
         'alignment_view' => 0,
       }
 
+      form.merge!(myspecies) if myspecies
+
       form.keys.each do |k|
         form.delete(k) unless form[k]
       end
@@ -227,10 +238,9 @@ module Bio::Blast::Remote
           end
         end
 
-        # workaround 2005.08.12
-        if /\<A +HREF=\"(http\:\/\/blast\.genome\.jp(\/tmp\/[^\"]+))\"\>Show all result\<\/A\>/i =~ @output.to_s then
-          result = http.get($2)
-          @output = result.body
+        # workaround 2005.08.12 + 2011.01.27
+        if /\<A +HREF=\"(http\:\/\/[\-\.a-z0-9]+\.genome\.jp(\/tmp\/[^\"]+))\"\>Show all result\<\/A\>/i =~ @output.to_s then
+          @output = Bio::Command.read_uri($1)
           txt = @output.to_s.split(/\<pre\>/)[1]
           raise 'cannot understand response' unless txt
           txt.sub!(/\<\/pre\>.*\z/m, '')

data/lib/bio/db/embl/sptr.rb

@@ -50,7 +50,18 @@ class SPTR < EMBLDB
   # returns a content (Int or String) of the ID line by a given key.
   # Hash keys: ['ENTRY_NAME', 'DATA_CLASS', 'MODECULE_TYPE', 'SEQUENCE_LENGTH']
   #
-  # === ID Line
+  # === ID Line (since UniProtKB release 9.0 of 31-Oct-2006)
+  #   ID   P53_HUMAN               Reviewed;         393 AA.
+  #   #"ID  #{ENTRY_NAME} #{DATA_CLASS}; #{SEQUENCE_LENGTH}."
+  #
+  # === Examples
+  #   obj.id_line  #=> {"ENTRY_NAME"=>"P53_HUMAN", "DATA_CLASS"=>"Reviewed", 
+  #                     "SEQUENCE_LENGTH"=>393, "MOLECULE_TYPE"=>nil}
+  #
+  #   obj.id_line('ENTRY_NAME') #=> "P53_HUMAN"
+  #
+  # 
+  # === ID Line (older style)
   #   ID   P53_HUMAN      STANDARD;      PRT;   393 AA.
   #   #"ID  #{ENTRY_NAME} #{DATA_CLASS}; #{MOLECULE_TYPE}; #{SEQUENCE_LENGTH}."
   #
@@ -65,11 +76,20 @@ class SPTR < EMBLDB
     return @data['ID'] if @data['ID']
 
     part = @orig['ID'].split(/ +/)         
+    if part[4].to_s.chomp == 'AA.' then
+      # after UniProtKB release 9.0 of 31-Oct-2006
+      # (http://www.uniprot.org/docs/sp_news.htm)
+      molecule_type   = nil
+      sequence_length = part[3].to_i
+    else
+      molecule_type   = part[3].sub(/;/,'')
+      sequence_length = part[4].to_i
+    end
     @data['ID'] = {
       'ENTRY_NAME'      => part[1],
       'DATA_CLASS'      => part[2].sub(/;/,''),
-      'MOLECULE_TYPE'   => part[3].sub(/;/,''),
-      'SEQUENCE_LENGTH' => part[4].to_i 
+      'MOLECULE_TYPE'   => molecule_type,
+      'SEQUENCE_LENGTH' => sequence_length
     }
   end
 
@@ -111,12 +131,27 @@ class SPTR < EMBLDB
   # returns a Hash of information in the DT lines.
   #  hash keys: 
   #    ['created', 'sequence', 'annotation']
+  #--
   #  also Symbols acceptable (ASAP):
   #    [:created, :sequence, :annotation]
+  #++
   #
-  # returns a String of information in the DT lines by a given key..
+  # Since UniProtKB release 7.0 of 07-Feb-2006, the DT line format is
+  # changed, and the word "annotation" is no longer used in DT lines.
+  # Despite the change, the word "annotation" is still used for keeping
+  # compatibility.
+  #
+  # returns a String of information in the DT lines by a given key.
   #
   # === DT Line; date (3/entry)
+  #   DT DD-MMM-YYY (integrated into UniProtKB/XXXXX.)
+  #   DT DD-MMM-YYY (sequence version NN)
+  #   DT DD-MMM-YYY (entry version NN)
+  #
+  # The format have been changed in UniProtKB release 7.0 of 07-Feb-2006.
+  # Below is the older format.
+  #
+  # === Old format of DT Line; date (3/entry)
   #   DT DD-MMM-YYY (rel. NN, Created)
   #   DT DD-MMM-YYY (rel. NN, Last sequence update)
   #   DT DD-MMM-YYY (rel. NN, Last annotation update)
@@ -133,7 +168,79 @@ class SPTR < EMBLDB
   end
 
 
+  # (private) parses DE line (description lines)
+  # since UniProtKB release 14.0 of 22-Jul-2008
+  #
+  # Return array containing array.
+  #
+  # http://www.uniprot.org/docs/sp_news.htm
+  def parse_DE_line_rel14(str)
+    # Retruns if it is not the new format since Rel.14
+    return nil unless /^DE   (RecName|AltName|SubName)\: / =~ str
+    ret = []
+    cur = nil
+    str.each_line do |line|
+      case line
+      when /^DE   (Includes|Contains)\: *$/
+        cur = [ $1 ]
+        ret.push cur
+        cur = nil
+        #subcat_and_desc = nil
+        next
+      when /^DE   *(RecName|AltName|SubName)\: +(.*)/
+        category = $1
+        subcat_and_desc = $2
+        cur = [ category ]
+        ret.push cur
+      when /^DE   *(Flags)\: +(.*)/
+        category = $1
+        desc = $2
+        flags = desc.strip.split(/\s*\;\s*/) || []
+        cur = [ category, flags ]
+        ret.push cur
+        cur = nil
+        #subcat_and_desc = nil
+        next
+      when /^DE   *(.*)/
+        subcat_and_desc = $1
+      else
+        warn "Warning: skipped DE line in unknown format: #{line.inspect}"
+        #subcat_and_desc = nil
+        next
+      end
+      case subcat_and_desc
+      when nil
+        # does nothing
+      when /\A([^\=]+)\=(.*)/
+        subcat = $1
+        desc = $2
+        desc.sub!(/\;\s*\z/, '')
+        unless cur
+          warn "Warning: unknown category in DE line: #{line.inspect}"
+          cur = [ '' ]
+          ret.push cur
+        end
+        cur.push [ subcat, desc ]
+      else
+        warn "Warning: skipped DE line description in unknown format: #{line.inspect}"
+      end
+    end
+    ret
+  end
+  private :parse_DE_line_rel14
+
   # returns the proposed official name of the protein.
+  # Returns a String.
+  #
+  # Since UniProtKB release 14.0 of 22-Jul-2008, the DE line format have
+  # been changed. The method returns the full name which is taken from
+  # "RecName: Full=" or "SubName: Full=" line normally in the beginning of
+  # the DE lines. 
+  # Unlike parser for old format, no special treatments for fragment or
+  # precursor.
+  #
+  # For old format, the method parses the DE lines and returns the protein
+  # name as a String.
   # 
   # === DE Line; description (>=1)
   #  "DE #{OFFICIAL_NAME} (#{SYNONYM})"
@@ -142,27 +249,83 @@ class SPTR < EMBLDB
   #  SYNONYM        >=0
   #  CONTEINS       >=0
   def protein_name
-    name = ""
-    if de_line = fetch('DE') then
-      str = de_line[/^[^\[]*/] # everything preceding the first [ (the "contains" part)
-      name = str[/^[^(]*/].strip
-      name << ' (Fragment)' if str =~ /fragment/i
+    @data['DE'] ||= parse_DE_line_rel14(get('DE'))
+    parsed_de_line = @data['DE']
+    if parsed_de_line then
+      # since UniProtKB release 14.0 of 22-Jul-2008
+      name = nil
+      parsed_de_line.each do |a|
+        case a[0]
+        when 'RecName', 'SubName'
+          if name_pair = a[1..-1].find { |b| b[0] == 'Full' } then
+            name = name_pair[1]
+            break
+          end
+        end
+      end
+      name = name.to_s
+    else
+      # old format (before Rel. 13.x)
+      name = ""
+      if de_line = fetch('DE') then
+        str = de_line[/^[^\[]*/] # everything preceding the first [ (the "contains" part)
+        name = str[/^[^(]*/].strip
+        name << ' (Fragment)' if str =~ /fragment/i
+      end
     end
     return name
   end
 
 
-  # returns an array of synonyms (unofficial names).
+  # returns synonyms (unofficial and/or alternative names).
+  # Returns an Array containing String objects.
+  #
+  # Since UniProtKB release 14.0 of 22-Jul-2008, the DE line format have
+  # been changed. The method returns the full or short names which are
+  # taken from "RecName: Short=", "RecName: EC=", and AltName lines,
+  # except after "Contains:" or "Includes:".
+  # For keeping compatibility with old format parser, "RecName: EC=N.N.N.N"
+  # is reported as "EC N.N.N.N".
+  # In addition, to prevent confusion, "Allergen=" and "CD_antigen=" 
+  # prefixes are added for the corresponding fields.
   #
+  # For old format, the method parses the DE lines and returns synonyms.
   # synonyms are each placed in () following the official name on the DE line.
   def synonyms
     ary = Array.new
-    if de_line = fetch('DE') then
-      line = de_line.sub(/\[.*\]/,'') # ignore stuff between [ and ].  That's the "contains" part
+    @data['DE'] ||= parse_DE_line_rel14(get('DE'))
+    parsed_de_line = @data['DE']
+    if parsed_de_line then
+      # since UniProtKB release 14.0 of 22-Jul-2008
+      parsed_de_line.each do |a|
+        case a[0]
+        when 'Includes', 'Contains'
+          break #the each loop
+        when 'RecName', 'SubName', 'AltName'
+          a[1..-1].each do |b|
+            if name = b[1] and b[1] != self.protein_name then
+              case b[0]
+              when 'EC'
+                name = "EC " + b[1]
+              when 'Allergen', 'CD_antigen'
+                name = b[0] + '=' + b[1]
+              else
+                name = b[1]
+              end
+              ary.push name
+            end
+          end
+        end #case a[0]
+      end #parsed_de_line.each
+    else
+      # old format (before Rel. 13.x)
+      if de_line = fetch('DE') then
+        line = de_line.sub(/\[.*\]/,'') # ignore stuff between [ and ].  That's the "contains" part
       line.scan(/\([^)]+/) do |synonym| 
         unless synonym =~ /fragment/i then 
           ary << synonym[1..-1].strip # index to remove the leading (  
         end
+        end
       end
     end
     return ary
@@ -919,25 +1082,34 @@ class SPTR < EMBLDB
   end
   private :cc_subcellular_location
 
-  
-  # CC   -!- WEB RESOURCE: NAME=ResourceName[; NOTE=FreeText][; URL=WWWAddress].  
+
+  #--
+  # Since UniProtKB release 12.2 of 11-Sep-2007:
+  # CC   -!- WEB RESOURCE: Name=ResourceName[; Note=FreeText][; URL=WWWAddress].  # Old format:
+  # CC   -!- WEB RESOURCE: NAME=ResourceName[; NOTE=FreeText][; URL=WWWAddress].
+  #++
+
   def cc_web_resource(data)
     data.map {|x|
-      entry = {'NAME' => nil, 'NOTE' => nil, 'URL' => nil}
+      entry = {'Name' => nil, 'Note' => nil, 'URL' => nil}
       x.split(';').each do |y|
         case y
-        when /NAME=(.+)/
-          entry['NAME'] = $1.strip
-        when /NOTE=(.+)/
-          entry['NOTE'] = $1.strip
-        when /URL="(.+)"/
+        when /(Name|Note)\=(.+)/
+          key = $1
+          val = $2.strip
+          entry[key] = val
+        when /(NAME|NOTE)\=(.+)/
+          key = $1.downcase.capitalize
+          val = $2.strip
+          entry[key] = val
+        when /URL\=\"(.+)\"/
           entry['URL'] = $1.strip
         end
       end
       entry
     }
   end
-  
+  private :cc_web_resource
 
   # returns databases cross-references in the DR lines.
   # * Bio::SPTR#dr  -> Hash w/in Array