protk 1.3.0 → 1.3.1.pre2

data/lib/protk.rb

@@ -1,7 +1,6 @@
 require 'protk/tool.rb'
 require 'protk/swissprot_database.rb'
 require 'protk/search_tool.rb'
-require 'protk/protxml.rb'
 require 'protk/prophet_tool.rb'
 require 'protk/omssa_util.rb'
 require 'protk/mascot_util.rb'

data/lib/protk/bio_gff3_extensions.rb

@@ -0,0 +1,22 @@
+require 'bio'
+
+# Extension to GFF3 records to support genomic coordinate mapping tasks
+
+class Bio::GFF::GFF3::Record
+
+
+# Comparator to allow sorting by start
+
+# Overlap operator to return a new gff by overlapping this one with another
+
+# Function to return our coordinates relative to some other coordinate system (eg a protein)
+
+	def <=>(otherRecord)
+		self.start <=> otherRecord.start
+	end
+
+	def length
+		return self.end-self.start+1
+	end
+
+end

data/lib/protk/bio_sptr_extensions.rb

@@ -90,7 +90,7 @@ class Bio::SPTR < Bio::EMBLDB
   # SwissProt Accessions
   #
   def accessions 
-    return ""
+    return self.ac
   end
 
   # Subcellular Location
@@ -132,7 +132,8 @@ class Bio::SPTR < Bio::EMBLDB
   def domain
     return self.cc["DOMAIN"].to_s
   end
-  
+
+
   # 
   # Getting dr entry
   # 
@@ -152,6 +153,20 @@ class Bio::SPTR < Bio::EMBLDB
   def ipi
     return self.safely_get_drentry_for_key("IPI")
   end
+
+  def go_terms
+    terms = self.dr["GO"]
+    if terms
+      return terms.collect { |e| e[0] }
+    else
+      return nil
+    end
+  end
+
+  def go_entries
+    return self.dr["GO"]
+  end  
+
   
   # Intact accession number
   #
@@ -234,8 +249,8 @@ class Bio::SPTR < Bio::EMBLDB
     return self.seq.to_s
   end
 
-  def tax_dump
-    return self.ox.to_s
+  def ncbi_taxon_id
+    return self.ox["NCBI_TaxID"]
   end
 
   def species_dump

data/lib/protk/constants.rb

@@ -29,16 +29,19 @@ class Constants
   attr :info_level
   attr :protk_dir
   attr :data_lib_dir
+  attr_accessor :info_level
 
   # Provides direct access to constants through methods of the same name
   # This will be used for all constants other than paths
   #
   def method_missing(method)
+
     from_env = @env[method.to_s]
     throw "#{method} is undefined" unless from_env!=nil
     from_env
   end
 
+
   # Some constants are paths. They need to be translated into real paths before being returned
   #
   
@@ -121,7 +124,7 @@ class Constants
   # Read the global constants file and initialize our class @env variable
   # Initialize loggers
   #
-  def initialize 
+  def initialize() 
 
     @data_lib_dir="#{File.dirname(__FILE__)}/data"
     @protk_dir="#{Dir.home}/.protk"
@@ -170,8 +173,10 @@ class Constants
 
     # puts "Path #{ENV['PATH']}"
     throw "No data found in config file" unless @env!=nil
-    @info_level=default_config_yml['message_level']
-    
+
+    @info_level="fatal"
+    @info_level=default_config_yml['message_level'] unless default_config_yml['message_level'].nil?
+
   end
 
 
@@ -196,15 +201,17 @@ class Constants
     throw "Unable to create file logger at path #{self.log_file}" unless @file_logger!=nil
     throw "Unable to create stdout logger " unless @stdout_logger!=nil
 
-  
-
     case @info_level
-    when "info"
+    when /info/i
       @stdout_logger.level=Logger::INFO
-    when "debug"    
+    when /debug/i    
       @stdout_logger.level=Logger::DEBUG
-    when "warn"
-      @stdout_logger.level=Logger::WARN      
+    when /warn/i
+      @stdout_logger.level=Logger::WARN
+    when /fatal/i
+      @stdout_logger.level=Logger::FATAL
+    else
+      throw "Unknown log level #{@info_level}"
     end
 
   end
@@ -215,8 +222,9 @@ class Constants
     if ( @stdout_logger == nil || @file_logger == nil)
       initialize_loggers
     end
-    @stdout_logger.send(level,message)
-    @file_logger.send(level,message)        
+
+   @stdout_logger.send(level,message)
+   @file_logger.send(level,message)        
   end
 
   def path_for_builtin_database(dbname)

data/lib/protk/gffdb.rb

@@ -0,0 +1,60 @@
+require 'protk/constants'
+require 'bio'
+
+
+class GFFDB
+
+  attr_accessor :id_to_records_map
+
+  def initialize(gff_file_path)
+    env = Constants.new
+    @database = gff_file_path
+    @id_to_records_map={}
+    @id_to_cds_map={}
+  end
+
+  def self.create(gff_file_path)
+    db = GFFDB.new(gff_file_path)
+    db.make_index(gff_file_path)
+    db
+  end
+
+  def get_by_id(entry_id)
+    @id_to_records_map[entry_id]
+  end
+
+  def get_cds_by_parent_id(entry_id)
+    @id_to_cds_map[entry_id]
+  end
+
+
+  def make_index(input_gff)
+    io = File.open(input_gff, "r")
+    gffdb = Bio::GFF::GFF3.new(io)  #parses the entire db
+
+    # Now create the mapping from ids to records
+    gffdb.records.each do |record| 
+
+      @id_to_records_map[record.id] = [] if @id_to_records_map[record.id].nil?
+      @id_to_records_map[record.id] << record
+
+      begin
+        # puts record.feature_type.match(/CDS/)
+        if record.feature_type.to_s =~ /CDS/i
+          # puts record.feature_type
+          parent_id=record.attributes_to_hash['Parent']
+          # puts parent_id
+          if parent_id
+            @id_to_cds_map[parent_id] = [] if @id_to_cds_map[parent_id].nil?
+            @id_to_cds_map[parent_id] << record
+          end
+        end
+
+      rescue
+        puts "Problem initializing cds map for #{record}"
+      end
+    end
+
+  end
+
+end

data/lib/protk/peptide.rb

@@ -0,0 +1,158 @@
+require 'libxml'
+require 'bio'
+require 'protk/bio_gff3_extensions'
+include LibXML
+
+class PeptideNotInProteinError < StandardError
+end
+
+class Peptide
+
+	attr_accessor :sequence
+	attr_accessor :protein_name
+	attr_accessor :charge
+	attr_accessor :nsp_adjusted_probability
+
+
+
+	class << self
+		def from_protxml(xmlnode)
+			pep=new()
+			pep.sequence=xmlnode['peptide_sequence']
+			pep.nsp_adjusted_probability=xmlnode['nsp_adjusted_probability'].to_f
+			pep.charge=xmlnode['charge'].to_i
+			pep
+		end
+
+		def from_sequence(seq,charge=nil)
+			pep=new()
+			pep.sequence=seq
+			pep.charge=charge
+			pep
+		end
+		private :new
+	end
+
+	def initialize()
+
+	end
+
+	# Expects prot_seq not to contain explicit stop codon (ie * at end)
+	# AA coords are 0-based unlike genomic coords which are 1 based
+	#
+	def coords_in_protein(prot_seq,reverse=false)
+		if reverse
+			pep_index = prot_seq.reverse.index(self.sequence.reverse)
+			raise PeptideNotInProteinError if pep_index.nil?
+			pep_start_i = pep_index
+		else
+			pep_start_i = prot_seq.index(self.sequence)
+			raise PeptideNotInProteinError if pep_start_i.nil?			
+		end
+		pep_end_i = pep_start_i+self.sequence.length
+		{:start => pep_start_i,:end => pep_end_i}
+	end
+
+
+	# Returns a list of fragments (hashes with start and end) in GFF style (1 based) genomic coordinates
+	#
+	# Assumes that cds_coords is inclusive of the entire protein sequence including start-met
+	#
+	# We assume that gff records conform to the spec
+	#
+	# http://www.sequenceontology.org/gff3.shtml
+	#
+	# This part of the spec is crucial
+	#
+	# - The START and STOP codons are included in the CDS. 
+	# - That is, if the locations of the start and stop codons are known, 
+	# - the first three base pairs of the CDS should correspond to the start codon
+	# - and the last three correspond the stop codon.
+	#
+	# We also assume that all the cds records provided, actually form part of the protein (ie skipped exons should not be included)
+	#
+	def to_gff3_records(prot_seq,parent_record,cds_records)
+
+		throw "Expected GFF3 Record but got #{parent_record.class}" unless parent_record.class==Bio::GFF::GFF3::Record
+		throw "Expected Array but got #{cds_records.class}" unless cds_records.class==Array
+
+		on_reverse_strand = (parent_record.strand=="-") ? true : false
+		aa_coords = coords_in_protein(prot_seq,false) # Always use forward protein coordinates
+
+		ordered_cds_records = on_reverse_strand ? cds_records.sort.reverse : cds_records.sort
+
+		# Initial position is the number of NA's from the start of translation
+		#
+		pep_nalen = self.sequence.length*3
+
+		i = 0; #Current protein position (in nucleic acids)
+
+		pep_start_i = aa_coords[:start]*3
+		pep_end_i = pep_start_i+self.sequence.length*3
+		fragments=[]
+		ordered_cds_records.each do |cds_record|
+			# puts cds_record
+			fragment = nil
+			fragment_len = 0
+			if on_reverse_strand
+
+				in_peptide = (i<pep_end_i) && (i>=pep_start_i)
+				before_len = [pep_start_i-i,0].max
+				# puts before_len
+				# puts in_peptide
+				# puts "i #{i} pi #{pep_end_i} psi #{pep_start_i}"
+				if in_peptide
+
+					fragment_end = cds_record.end
+					fragment_len = [cds_record.length,pep_end_i-i].min
+					fragment_start = fragment_end-fragment_len+1
+					# fragment = {:start=>fragment_start,:end=>fragment_end}
+					fragment = gff_record_for_peptide_fragment(fragment_start,fragment_end,cds_record)
+
+				elsif before_len>0
+					fragment_end = cds_record.end - before_len
+					fragment_len = [cds_record.length-before_len,pep_end_i-i-before_len].min
+					# puts "Frag len #{fragment_len}"
+					fragment_start = fragment_end - fragment_len + 1
+					fragment = gff_record_for_peptide_fragment(fragment_start,fragment_end,cds_record)
+					# fragment = {:start=>fragment_start,:end=>fragment_end}
+				else
+					fragment=nil
+				end				
+			else
+				in_peptide = (i<pep_end_i) && (i>=pep_start_i)
+				before_len = [pep_start_i-i,0].max
+				if in_peptide
+					fragment_start = cds_record.start
+					fragment_len = [cds_record.length,pep_end_i-i].min
+					fragment_end = fragment_start+fragment_len-1
+					# fragment = {:start=>fragment_start,:end=>fragment_end}
+					fragment = gff_record_for_peptide_fragment(fragment_start,fragment_end,cds_record)
+				elsif before_len>0
+					fragment_start = cds_record.start + before_len
+					fragment_len = [cds_record.length-before_len,pep_end_i-i-before_len].min
+					fragment_end = fragment_start + fragment_len-1
+					# fragment = {:start=>fragment_start,:end=>fragment_end}
+					fragment = gff_record_for_peptide_fragment(fragment_start,fragment_end,cds_record)
+				else
+					fragment=nil
+				end
+
+			end
+			i+=cds_record.length
+			fragments << fragment unless fragment.nil?
+		end
+		fragments
+	end
+
+	def gff_record_for_peptide_fragment(start_i,end_i,parent_record)
+		cds_id = parent_record.id
+		this_id = "#{cds_id}.#{self.sequence}"
+		this_id << ".#{self.charge}" unless self.charge.nil?
+		score = self.nsp_adjusted_probability.nil? ? "." : self.nsp_adjusted_probability.to_s
+		gff_string = "#{parent_record.seqid}\tMSMS\tpolypeptide\t#{start_i}\t#{end_i}\t#{score}\t#{parent_record.strand}\t0\tID=#{this_id};Parent=#{cds_id}"
+		Bio::GFF::GFF3::Record.new(gff_string)
+	end
+
+
+end

data/lib/protk/protein.rb

@@ -0,0 +1,72 @@
+require 'protk/peptide'
+
+include LibXML
+
+
+class Protein
+
+	attr_accessor :group_number
+	attr_accessor :group_probability
+	attr_accessor :probability
+	attr_accessor :sequence
+	attr_accessor :protein_name
+	attr_accessor :n_indistinguishable_proteins
+	attr_accessor :percent_coverage
+	attr_accessor :peptides
+
+	class << self
+
+		# <protein_group group_number="1" probability="1.0000">
+		#       <protein protein_name="ACADV_MOUSE" n_indistinguishable_proteins="1" probability="1.0000" percent_coverage="9.9" unique_stripped_peptides="ELGAFGLQVPSELGGLGLSNTQYAR+GIVNEQFLLQR+SGELAVQALDQFATVVEAK+VAVNILNNGR" group_sibling_id="a" total_number_peptides="4" pct_spectrum_ids="0.41" confidence="1.00">
+		#          <parameter name="prot_length" value="656"/>
+		#          <annotation protein_description="Very long-chain specific acyl-CoA dehydrogenase, mitochondrial OS=Mus musculus GN=Acadvl PE=1 SV=3"/>
+		#          <peptide peptide_sequence="SGELAVQALDQFATVVEAK" charge="1" initial_probability="0.9919" nsp_adjusted_probability="0.9981" weight="1.00" is_nondegenerate_evidence="Y" n_enzymatic_termini="2" n_sibling_peptides="2.34" n_sibling_peptides_bin="5" n_instances="1" exp_tot_instances="0.99" is_contributing_evidence="Y" calc_neutral_pep_mass="1975.0340">
+		#          </peptide>
+		#          <peptide peptide_sequence="GIVNEQFLLQR" charge="1" initial_probability="0.9909" nsp_adjusted_probability="0.9979" weight="1.00" is_nondegenerate_evidence="Y" n_enzymatic_termini="2" n_sibling_peptides="2.34" n_sibling_peptides_bin="5" n_instances="1" exp_tot_instances="0.99" is_contributing_evidence="Y" calc_neutral_pep_mass="1315.7250">
+		#          </peptide>
+		#          <peptide peptide_sequence="ELGAFGLQVPSELGGLGLSNTQYAR" charge="1" initial_probability="0.7792" nsp_adjusted_probability="0.9391" weight="1.00" is_nondegenerate_evidence="Y" n_enzymatic_termini="2" n_sibling_peptides="2.55" n_sibling_peptides_bin="5" n_instances="1" exp_tot_instances="0.78" is_contributing_evidence="Y" calc_neutral_pep_mass="2576.3234">
+		#          </peptide>
+		#          <peptide peptide_sequence="VAVNILNNGR" charge="1" initial_probability="0.5674" nsp_adjusted_probability="0.8515" weight="1.00" is_nondegenerate_evidence="Y" n_enzymatic_termini="2" n_sibling_peptides="2.76" n_sibling_peptides_bin="5" n_instances="1" exp_tot_instances="0.57" is_contributing_evidence="Y" calc_neutral_pep_mass="1068.6030">
+		#          </peptide>
+		#       </protein>
+		# </protein_group>
+
+
+		def from_protxml(xmlnode)
+			prot=new()
+			groupnode = xmlnode.parent
+			prot.group_probability = groupnode['probability'].to_f
+			prot.group_number = groupnode['group_number'].to_i
+			prot.probability = xmlnode['probability'].to_f
+			prot.protein_name = xmlnode['protein_name']
+			prot.n_indistinguishable_proteins = xmlnode['n_indistinguishable_proteins'].to_i
+			prot.percent_coverage = xmlnode['percent_coverage'].to_f
+
+			peptide_nodes = xmlnode.find('protxml:peptide','protxml:http://regis-web.systemsbiology.net/protXML')
+			prot.peptides = peptide_nodes.collect { |e| Peptide.from_protxml(e) }
+			prot
+		end
+		private :new
+	end
+
+	def initialize()
+
+	end
+
+	# Return just one peptide for each unique sequence choosing the peptide with highest probability
+	#
+	def representative_peptides()
+		best_peptides={}
+		self.peptides.each do |peptide|
+			seq = peptide.sequence
+			if best_peptides[seq].nil?
+				best_peptides[seq]=peptide				
+			else
+				best_peptides[seq]=peptide if peptide.nsp_adjusted_probability > best_peptides[seq].nsp_adjusted_probability
+			end
+		end
+
+		best_peptides.values
+	end
+
+end