mspire 0.1.5 → 0.1.7

data/lib/spec_id/sequest.rb

@@ -6,6 +6,74 @@ require 'set_from_hash'
 require 'spec_id/bioworks'
 require 'instance_var_set_from_hash'
 require 'spec/msrun'
+require 'spec_id/srf'
+
+class Numeric
+  # returns a string with a + or - on the front
+  def to_plus_minus_string
+    if self >= 0
+      '+' << self.to_s
+    else
+      '-' << self.to_s
+    end
+  end
+end
+
+##########################################
+# NEED TO ADD MODIFICATIONS and generally verify pepxml creation!!! :
+# HERE's an excerpt from an example file from tpp 2.9.2 that I'm going to follow:
+=begin
+<search_summary base_name="/regis/data3/search/akeller/LCQ/COMET/LIGHT/haloICAT2_41" search_engine="COMET" precursor_mass_type="average" fragment_mass_type="average">
+<sequence_search_constraint sequence="C"/>
+<aminoacid_modification aminoacid="C" massdiff="8.049" mass="553.765" variable="Y" binary="N"/>
+<aminoacid_modification aminoacid="C" massdiff="442.5772" mass="545.7160" variable="N"/>
+<aminoacid_modification aminoacid="M" massdiff="16.0000" mass="147.1926" variable="Y" binary="N" symbol="1"/>
+<parameter name="peptide_mass_tol" value="3.0000"/>
+<parameter name="peptide_mass_tol_units" value="DA"/>
+<parameter name="num_output_lines" value="10"/>
+<parameter name="remove_precursor_peak" value="0"/>
+<parameter name="num_dup_headers" value="1"/>
+<parameter name="email_address" value=""/>
+<parameter name="ion_series" value="010000010"/>
+<parameter name="max_num_var_mod_residues" value="3"/>
+<parameter name="md5_check_sum" value="2547286a77a35abe2af3f2e9825ab814"/>
+</search_summary>
+=end
+
+# and a guy with modifications:
+=begin
+<search_result spectrum="haloICAT2_41.1110.1110.2" start_scan="1110" end_scan="1110" precursor_neutral_mass="2000.6641" assumed_charge="2" index="28">
+<search_hit hit_rank="1" peptide="GCMPSKEVLSAGAHR" peptide_prev_aa="R" peptide_next_aa="Y" protein="Chr_ORF0132" num_tot_proteins="1" num_matched_ions="19" tot_num_ions="30" calc_neutral_pep_mass="2001.3685" massdiff="-0.704" num_tol_term="2" num_missed_cleavages="1" is_rejected="0">
+<modification_info modified_peptide="GC[546]M[147]PSKEVLSAGAHR">
+<mod_aminoacid_mass position="2" mass="545.7160"/>
+<mod_aminoacid_mass position="3" mass="147.1926"/>
+</modification_info>
+<search_score name="dotproduct" value="359"/>
+<search_score name="delta" value="0.296"/>
+<search_score name="deltastar" value="0"/>
+<search_score name="zscore" value="5.290"/>
+<search_score name="expect" value="0.000E+00"/>
+<peptideprophet_result probability="0.9994" all_ntt_prob="(0.3713,0.4360,0.9994)">
+<search_score_summary>
+<parameter name="fval" value="3.4002"/>
+<parameter name="ntt" value="2"/>
+<parameter name="nmc" value="1"/>
+<parameter name="massd" value="-0.704"/>
+</search_score_summary>
+</peptideprophet_result>
+=end
+
+# sequest.params option:
+# diff_search_options = 15.994910 M 0.000000 C 0.000000 M 0.000000 X 0.000000 T 0.000000 Y
+# permanent mods are at the bottom: ...
+# add_A_Alanine = 0.0000                   ; added to A
+# add_S_Serine = 0.0000                    ; added to S
+# add_P_Proline = 0.0000                   ; added to P
+# add_V_Valine = 0.0000                    ; added to V
+# add_T_Threonine = 0.0000                 ; added to T
+# ...
+
+
 
 module SpecID::Sequest; end
 class SpecID::Sequest::PepXML; end
@@ -26,8 +94,12 @@ class SpecID::Sequest::PepXML::MSMSPipelineAnalysis
     @xmlns = nil
     @xmlns_xsi = nil
     @xsi_schema_location = nil
-    self.set_from_hash(hash)
-    @msms_run_summary = yield
+    if hash
+      self.set_from_hash(hash)
+    end
+    if block_given?
+      @msms_run_summary = yield
+    end
   end
 
   # if no date string given, then it will set to Time.now
@@ -80,7 +152,8 @@ class SpecID::Sequest::PepXML::MSMSRunSummary
   include SpecIDXML
 
   # the version of TPP you are using (determines xml output)
-  # The name of the pep xml file (without extension)
+  # The name of the pep xml file (without extension) (but this is a long
+  # filename!!!)
   attr_accessor :base_name
   # The name of the mass spec manufacturer 
   attr_accessor :ms_manufacturer
@@ -104,7 +177,9 @@ class SpecID::Sequest::PepXML::MSMSRunSummary
   # set to the return value of the block
   def initialize(hash=nil)
     @spectrum_queries = []
-    instance_var_set_from_hash(hash)
+    if hash
+      instance_var_set_from_hash(hash)
+    end
     if block_given? ; @spectrum_queries = yield end
   end
 
@@ -137,6 +212,8 @@ end
 
 class SpecID::Sequest::PepXML
   include SpecIDXML
+
+  ## CREATE a default version for the entire class
   class << self
     attr_accessor :pepxml_version
   end
@@ -144,7 +221,11 @@ class SpecID::Sequest::PepXML
   self.pepxml_version = DEF_VERSION # default version
 
   attr_accessor :pepxml_version, :msms_pipeline_analysis
+  ## the full path name (no extension)
   attr_accessor :base_name
+  attr_accessor :h_plus
+  attr_accessor :avg_parent
+
   #attr_accessor :spectrum_queries, :params, :base_name, :search_engine, :database, :raw_data_type, :raw_data, :out_data_type, :out_data, :sample_enzyme, :pepxml_version
 
   # returns an array of spectrum queries
@@ -153,10 +234,29 @@ class SpecID::Sequest::PepXML
   end
 
   # msms_pipeline_analysis is set to the result of the yielded block
-  def initialize(pepxml_version=DEF_VERSION)
+  # and set_mono_or_avg is called with params if given
+  def initialize(pepxml_version=DEF_VERSION, sequest_params_obj=nil)
     self.class.pepxml_version = pepxml_version
-    @msms_pipeline_analysis = yield
-    @base_name = @msms_pipeline_analysis.msms_run_summary.base_name
+    if sequest_params_obj
+      set_mono_or_avg(sequest_params_obj)
+    end
+    if block_given?
+      @msms_pipeline_analysis = yield
+      @base_name = @msms_pipeline_analysis.msms_run_summary.base_name
+    end
+  end
+
+  # sets @h_plus and @avg_parent from the sequest params object
+  def set_mono_or_avg(sequest_params_obj)
+    case sequest_params_obj.precursor_mass_type
+    when "monoisotopic" ; @avg_parent = false
+    else ; @avg_parent = true
+    end
+
+    case @avg_parent
+    when true ; @h_plus = SpecID::AVG[:h_plus]
+    when false ; @h_plus = SpecID::MONO[:h_plus]
+    end
   end
 
   def date
@@ -203,6 +303,190 @@ class SpecID::Sequest::PepXML
     end
   end
 
+
+  Default_Options = {
+    :out_path => nil,
+    :backup_db_path => '/project/marcotte/marcotte/ms/database',
+    # a PepXML option
+    :pepxml_version => DEF_VERSION,  
+    ## MSMSRunSummary options:
+    # string must be recognized in sample_enzyme.rb 
+    # or create your own SampleEnzyme object
+    :sample_enzyme => 'trypsin',
+    :ms_manufacturer => 'ThermoFinnigan',
+    :ms_model => 'LCQ Deca XP',
+    :ms_ionization => 'ESI',
+    :ms_mass_analyzer => 'Ion Trap',
+    :ms_detector => 'UNKNOWN',
+    :raw_data_type => "raw",
+    :raw_data => ".mzXML", ## even if you don't have it?
+    ## SearchSummary options:
+    :out_data_type => "out", ## may be srf?? don't think pepxml recognizes this yet
+    :out_data => ".tgz" ## may be srf??
+  }
+
+  # will dynamically set :ms_model and :ms_mass_analyzer from srf info
+  # (ignoring defaults or anything passed in) for LTQ Orbitrap
+  # and LCQ Deca XP
+  # See SRF::Sequest::PepXML::Default_Options hash for defaults
+  # unless given, the out_path will be given as the path of the srf_file
+  def self.new_from_srf(srf_file, opts={})
+    opts = Default_Options.merge(opts)
+
+    ## set the outpath
+    out_path = opts.delete(:out_path)
+    unless out_path
+      out_path = File.dirname(srf_file)
+    end
+
+    ## read the srf file
+    srf = SRF.new(srf_file)
+
+    params = srf.params
+
+    ## check to see if we need backup_db
+    backup_db_path = opts.delete(:backup_db_path)
+    unless File.exist? params.database
+      params.database_path = backup_db_path
+    end
+
+    #######################################################################
+    # PREPARE THE OPTIONS:
+    #######################################################################
+    ## remove items from the options hash that don't belong to 
+    ppxml_version = opts.delete(:pepxml_version)
+    out_data_type = opts.delete(:out_data_type)
+    out_data = opts.delete(:out_data)
+
+    ## Extract meta info from srf
+    bn_noext = base_name_noext(srf.header.raw_filename)
+    opts[:ms_model] = srf.header.model
+    case opts[:ms_model]
+    when /Orbitrap/
+      opts[:ms_mass_analyzer] = 'Orbitrap'
+    when /LCQ Deca XP/
+      opts[:ms_mass_analyzer] = 'Ion Trap'
+    end
+
+    ## Create the base name
+    full_base_name_no_ext = make_base_name( File.expand_path(out_path), bn_noext)
+    opts[:base_name] = full_base_name_no_ext
+
+    ## Create the search summary:
+    search_summary_options = {
+      :search_database => SpecID::Sequest::PepXML::SearchDatabase.new(params),
+      :base_name => full_base_name_no_ext,
+      :out_data_type => out_data_type,
+      :out_data => out_data
+    }
+    opts[:search_summary] = SpecID::Sequest::PepXML::SearchSummary.new( params, search_summary_options)
+
+    ## Create the SampleEnzyme object if necessary
+    unless opts[:sample_enzyme].is_a? SampleEnzyme
+      opts[:sample_enzyme] = SampleEnzyme.new(opts[:sample_enzyme])
+    end
+
+    ## Create the pepxml obj
+    pepxml_obj = SpecID::Sequest::PepXML.new(ppxml_version, params) 
+    ## name some common variables we'll need
+    h_plus = pepxml_obj.h_plus
+    avg_parent = pepxml_obj.avg_parent
+
+    #######################################################################
+    # CREATE the spectrum_queries_ar
+    #######################################################################
+    srf_index = srf.index
+    out_files = srf.out_files
+    spectrum_queries_arr = Array.new(srf.dta_files.size)
+    files_with_hits_index = 0  ## will end up being 1 indexed
+    srf.dta_files.each_with_index do |dta_file,i|
+      next if out_files[i].num_hits == 0
+      files_with_hits_index += 1
+
+      # Sort the hits
+      hits = out_files[i].hits
+      arr = hits.sort_by{|v| v.xcorr }
+
+      # Get proper deltacn and deltacnstar
+      # Prophet deltacn is not the same as the native Sequest deltacn
+      # It is the deltacn of the second best hit!
+      top_hit = arr.pop
+      second_hit = arr.last
+      if second_hit 
+        top_hit[1] = second_hit[1]
+        deltacnstar = '0'
+      else 
+        top_hit[1] = '1.0'
+        deltacnstar = '1'
+      end
+
+      ## mass calculations:
+      precursor_neutral_mass = dta_file.mh - h_plus
+      calc_neutral_pep_mass = top_hit[0] - h_plus
+      massdiff = precursor_neutral_mass - calc_neutral_pep_mass
+      if massdiff >= 0 ; massdiff = "+" + massdiff.to_s
+      else ; massdiff = massdiff.to_s end
+
+      (start_scan, end_scan, charge) = srf_index[i]
+      sq_hash = {
+        :spectrum => [bn_noext, start_scan, end_scan, charge].join('.'),
+        :start_scan => start_scan,
+        :end_scan => end_scan,
+        :precursor_neutral_mass => precursor_neutral_mass,
+        :assumed_charge => charge,
+        :pepxml_version => ppxml_version,
+        :index => files_with_hits_index,
+      }
+
+      #  NEED TO MODIFY SPLIT SEQUENCE TO DO MODS!
+      ## THIS IS ALL INNER LOOP, so we make every effort at speed here:
+      (prevaa, pepseq, nextaa) = SpecID::Sequest::PepXML::SearchHit.prepare_sequence(top_hit[8])
+      #  ind_keys = {:mh => 0, :deltacn => 1, :sp => 2, :xcorr => 3, :id => 4, :rsp => 5, :ions_matched => 6, :ions_total => 7, :peptide => 8, :reference => 9 }
+
+      sh_hash = {
+        :hit_rank => "1",
+        :peptide => pepseq,
+        :peptide_prev_aa => prevaa,
+        :peptide_next_aa => nextaa,
+        :protein => top_hit[9].split(" ").first, 
+        :num_tot_proteins => top_hit[10],
+        :num_matched_ions => top_hit[6],
+        :tot_num_ions => top_hit[7],
+        :calc_neutral_pep_mass => calc_neutral_pep_mass,
+        :massdiff => massdiff, 
+        :num_tol_term => SpecID::Sequest::PepXML::SearchHit.calc_num_tol_term(params, top_hit[8]),
+        :num_missed_cleavages => SpecID::Sequest::PepXML::SearchHit.calc_num_missed_cleavages(params, top_hit[8]),
+        :is_rejected => '0',
+        # These are search score attributes:
+        :xcorr => top_hit[3],
+        :deltacn => top_hit[1],
+        :deltacnstar => deltacnstar,
+        :spscore => top_hit[2],
+        :sprank => top_hit[5],
+      }
+
+      spectrum_queries_arr[files_with_hits_index] = SpecID::Sequest::PepXML::SpectrumQuery.new(sq_hash) do
+        search_result = SpecID::Sequest::PepXML::SearchResult.new do
+          [ SpecID::Sequest::PepXML::SearchHit.new(sh_hash) ] # there can be multiple hits
+        end # SearchResult
+        [search_result] # can be multiple
+      end
+    end
+    spectrum_queries_arr.compact!
+
+    #######################################################################
+    # ADD the pipeline analysis
+    #######################################################################
+
+    pipeline = SpecID::Sequest::PepXML::MSMSPipelineAnalysis.new({:date=>nil,:summary_xml=> bn_noext +'.xml'}) do
+      SpecID::Sequest::PepXML::MSMSRunSummary.new(opts) { spectrum_queries_arr }
+    end
+    pepxml_obj.msms_pipeline_analysis = pipeline
+    pepxml_obj.base_name = pipeline.msms_run_summary.base_name
+    pepxml_obj
+  end
+
+  # Takes bioworks 3.2/3.3 xml output (with no filters)
   # Returns a list of PepXML objects
   # msdata = path to mzXML files (or .timeIndex files) (or @TODO: path to sqt file(s))
   # params = sequest.params file
@@ -246,6 +530,7 @@ class SpecID::Sequest::PepXML
 
     ## Create a hash of spectrum_query arrays by filename (this very big block):
     spectrum_queries_by_base_name = {}
+    pepxml_objs_by_base_name = {}
     # Hash by the filenames to split into filenames:
     bioworks.peps.hash_by(:base_name).each do |base_name, pep_arr|
 
@@ -262,7 +547,10 @@ class SpecID::Sequest::PepXML
         abort "invalid BioworksBrowser version: #{x}"
       end
 
-      spectrum_queries = pep_arr.hash_by(:first_scan, :last_scan, :charge).collect do |key,arr|
+      pepxml_obj = SpecID::Sequest::PepXML.new(pepxml_version, params)
+      pepxml_objs_by_base_name[base_name] = pepxml_obj
+
+      spectrum_queries_ar = pep_arr.hash_by(:first_scan, :last_scan, :charge).collect do |key,arr|
 
 
         # Sort_by_rank and take the top hit (to mimick out2summary):
@@ -270,28 +558,18 @@ class SpecID::Sequest::PepXML
         top_pep = arr.pop
         second_hit = arr.last # needed for deltacnstar
 
-        case params.precursor_mass_type
-        when "monoisotopic" ; avg_parent = false
-        else ; avg_parent = true
-        end
-
-        case avg_parent
-        when true ; h_plus = SpecID::AVG[:h_plus]
-        when false ; h_plus = SpecID::MONO[:h_plus]
-        end
 
-        
         case calc_prec_by
         when :prec_mz_arr
-          precursor_neutral_mass = SpecID::Sequest::PepXML::SpectrumQuery.calc_precursor_neutral_mass(calc_prec_by, top_pep.first_scan.to_i, top_pep.last_scan.to_i, prec_mz_arr, top_pep.charge.to_i, avg_parent)
+          precursor_neutral_mass = SpecID::Sequest::PepXML::SpectrumQuery.calc_precursor_neutral_mass(calc_prec_by, top_pep.first_scan.to_i, top_pep.last_scan.to_i, prec_mz_arr, top_pep.charge.to_i, pepxml_obj.avg_parent)
         when :deltamass
-          precursor_neutral_mass = SpecID::Sequest::PepXML::SpectrumQuery.calc_precursor_neutral_mass(calc_prec_by, top_pep.mass.to_f, top_pep.deltamass.to_f, avg_parent)
+          precursor_neutral_mass = SpecID::Sequest::PepXML::SpectrumQuery.calc_precursor_neutral_mass(calc_prec_by, top_pep.mass.to_f, top_pep.deltamass.to_f, pepxml_obj.avg_parent)
         end
 
-        calc_neutral_pep_mass = (top_pep.mass.to_f - h_plus)
+        calc_neutral_pep_mass = (top_pep.mass.to_f - pepxml_obj.h_plus)
         massdiff = precursor_neutral_mass - calc_neutral_pep_mass
         if massdiff >= 0 ; massdiff = "+" + massdiff.to_s
-        else ; massdiff = massdiff.to_s end
+        else ; massdiff = massdiff.to_s end #already has a -
         # deltacn & star:
         # (NOTE: OLD?? out2summary wants the deltacn of the 2nd best hit.)
         if second_hit 
@@ -317,7 +595,7 @@ class SpecID::Sequest::PepXML
             # NOTE: the bioworks mass is really M+H if two or more scans went
             # into the search_hit; calc_neutral_pep_mass is simply the avg of
             # precursor masses adjusted to be neutral
-            (prevaa, pepseq, nextaa) = SpecID::Sequest::PepXML::SearchHit.split_sequence(top_pep.sequence)
+            (prevaa, pepseq, nextaa) = SpecID::Sequest::PepXML::SearchHit.prepare_sequence(top_pep.sequence)
             (num_matched_ions, tot_num_ions) = SpecID::Sequest::PepXML::SearchHit.split_ions(top_pep.ions)
             search_hit = SpecID::Sequest::PepXML::SearchHit.new({
               :hit_rank => "1",
@@ -348,32 +626,36 @@ class SpecID::Sequest::PepXML
 
       # create an index by spectrum as results end up typically in out2summary
       # (I really dislike this order, however)
-      spectrum_queries = spectrum_queries.sort_by {|pep| pep.spectrum }
-      spectrum_queries.each_with_index {|res,index| res.index = "#{index + 1}" }
+      spectrum_queries_ar = spectrum_queries_ar.sort_by {|pep| pep.spectrum }
+      spectrum_queries_ar.each_with_index {|res,index| res.index = "#{index + 1}" }
 
-      spectrum_queries_by_base_name[base_name] = spectrum_queries
+      spectrum_queries_by_base_name[base_name] = spectrum_queries_ar
     end
 
-    spectrum_queries_by_base_name.collect do |base_name, spectrum_queries|
+    modifications_string = bioworks.modifications
+    
+    spectrum_queries_by_base_name.collect do |base_name, spectrum_queries_ar|
       case pepxml_version
       when 18
-        SpecID::Sequest::PepXML.new(pepxml_version) do
-          SpecID::Sequest::PepXML::MSMSPipelineAnalysis.new({:date=>nil,:summary_xml=>base_name+'.xml'}) do
-            full_base_name_no_ext = self.make_base_name( File.expand_path(out_path), base_name)
-            SpecID::Sequest::PepXML::MSMSRunSummary.new({
-              :base_name => full_base_name_no_ext,
-              :ms_manufacturer => ms_manufacturer,
-              :ms_model => ms_model,
-              :ms_ionization => ms_ionization,
-              :ms_mass_analyzer => ms_mass_analyzer,
-              :ms_detector => ms_detector,
-              :raw_data_type => raw_data_type,
-              :raw_data => raw_data,
-              :sample_enzyme => SampleEnzyme.new(sample_enzyme),
-              :search_summary => SpecID::Sequest::PepXML::SearchSummary.new(params, {:search_database => SpecID::Sequest::PepXML::SearchDatabase.new(params), :base_name => full_base_name_no_ext, :out_data_type => out_data_type, :out_data => out_data}),
-            }) do spectrum_queries end 
-          end
+        pipeline =  SpecID::Sequest::PepXML::MSMSPipelineAnalysis.new({:date=>nil,:summary_xml=>base_name+'.xml'}) do
+          full_base_name_no_ext = self.make_base_name( File.expand_path(out_path), base_name)
+          SpecID::Sequest::PepXML::MSMSRunSummary.new({
+            :base_name => full_base_name_no_ext,
+            :ms_manufacturer => ms_manufacturer,
+            :ms_model => ms_model,
+            :ms_ionization => ms_ionization,
+            :ms_mass_analyzer => ms_mass_analyzer,
+            :ms_detector => ms_detector,
+            :raw_data_type => raw_data_type,
+            :raw_data => raw_data,
+            :sample_enzyme => SampleEnzyme.new(sample_enzyme),
+            :search_summary => SpecID::Sequest::PepXML::SearchSummary.new(params, modifications_string, {:search_database => SpecID::Sequest::PepXML::SearchDatabase.new(params), :base_name => full_base_name_no_ext, :out_data_type => out_data_type, :out_data => out_data}),
+          }) { spectrum_queries_ar }
         end
+        pepxml_obj = pepxml_objs_by_base_name[base_name]
+        pepxml_obj.msms_pipeline_analysis = pipeline
+        pepxml_obj.base_name = pipeline.msms_run_summary.base_name
+        pepxml_obj 
       when 0
         ## @TODO: NEED TO REVAMP THIS:
         #        SpecID::Sequest::PepXML.new(pepxml_version).set_from_hash({
@@ -429,6 +711,14 @@ class SpecID::Sequest::PepXML
     string
   end
 
+  # given any kind of filename (from windows or whatever)
+  # returns the base of the filename with no file extension
+  def self.base_name_noext(file)
+    file.gsub!("\\", '/')
+    File.basename(file).sub(/\.[\w^\.]+$/, '')
+  end
+
+
 end # PepXML
 
 ##
@@ -461,6 +751,15 @@ class SpecID::Sequest::Params
         one,two = line.split @@param_re
         two,comment = two.split @@param_two_split
         hash[one] = two.rstrip
+        # it is necessary to add this break so that params files inside srf
+        # files can be read.  This will terminate the reading at the end of
+        # the file even though there are more lines
+        if line =~ /added to U/ || line =~ /digest_mass_range/## Will only work on bioworks 3.2 & 3.3 (bioworks 3.1 last line => Elastase/Tryp...)
+          break
+        end
+        if line =~ /digest_mass_range/  # there is no space in the srf params files
+          break
+        end
       else
         break
       end
@@ -468,17 +767,26 @@ class SpecID::Sequest::Params
     hash
   end
 
+  # returns self
+  def parse_handle(fh)
+    sequest_line = fh.gets #[SEQUEST]
+    @opts = grab_params(fh)
+    @opts["search_engine"] = "SEQUEST"
+    @mods = grab_params(fh)
+
+    ## this gets rid of the .hdr postfix on indexed databases
+    @opts["first_database_name"] = @opts["first_database_name"].sub(/\.hdr$/, '')
+    self
+  end
+
   ## parses file
   ## and drops the .hdr behind indexed fasta files
+  ## returns self
   def parse(file)
     File.open(file) do |fh|
-      sequest_line = fh.gets #[SEQUEST]
-      @opts = grab_params(fh)
-      @opts["search_engine"] = "SEQUEST"
-      @mods = grab_params(fh)
+      parse_handle(fh)
     end
-    ## this gets rid of the .hdr postfix on indexed databases
-    @opts["first_database_name"] = @opts["first_database_name"].sub(/\.hdr$/, '')
+    self
   end
 
   # returns( split_after, except_before)
@@ -569,6 +877,17 @@ class SpecID::Sequest::Params
     @opts["first_database_name"]
   end
 
+  # returns the appropriate aminoacid mass lookup table (in spec_id.rb SpecID::MONO or
+  # SpecID::AVG based on precursor_mass_type
+  def mass_table
+    case precursor_mass_type
+    when 'average'
+      SpecID::AVG
+    when 'monoisotopic'
+      SpecID::MONO
+    end
+  end
+
   # at least in Bioworks 3.2, the First number after the enzyme
   # is the indication of the enzymatic end stringency (required):
   #   1 = Fully enzymatic
@@ -628,7 +947,7 @@ class SpecID::Sequest::PepXML::SearchResult
   attr_accessor :search_hits
 
   # if block given, then search_hits set to return value
-  def initialize()
+  def initialize
     if block_given? ; @search_hits = yield
     else ; @search_hits = [] end
   end
@@ -646,13 +965,16 @@ class SpecID::Sequest::PepXML::SearchSummary
   attr_accessor :base_name
   attr_accessor :out_data_type
   attr_accessor :out_data
+  attr_accessor :modifications
   # A SearchDatabase object (responds to :local_path and :type)
   attr_accessor :search_database
   # if given a sequest params object, then will set the following attributes:
   # args is a hash of parameters
-  def initialize(params=nil, args=nil)
+  # modifications_string -> See Modifications
+  def initialize(params, modifications_string='', args=nil)
     @search_id = nil
     @params = params
+    @modifications = SpecID::Sequest::PepXML::Modifications.new(params, modifications_string)
     if args ; set_from_hash(args) end
   end
 
@@ -665,16 +987,304 @@ class SpecID::Sequest::PepXML::SearchSummary
     else ; '1' end
   end
 
+
   def to_pepxml
     element_xml(:search_summary, [:base_name, :search_engine, :precursor_mass_type, :fragment_mass_type, :out_data_type, :out_data, :search_id]) do
       search_database.to_pepxml +
         short_element_xml(:enzymatic_search_constraint, [:enzyme, :max_num_internal_cleavages, :min_number_termini]) +
+        @modifications.to_pepxml +
         @params.pepxml_parameters
     end
   end
 
 end
 
+class SpecID::Sequest::PepXML::Modifications
+  include SpecIDXML
+
+  # sequest params object
+  attr_accessor :params
+  # array holding AAModifications 
+  attr_accessor :aa_mods
+  # array holding TerminalModifications
+  attr_accessor :term_mods
+  # a hash of all differential modifications present by aa_one_letter_symbol
+  # and special_symbol. This is NOT the mass difference but the total mass {
+  # 'M*' => 155.5, 'S@' => 190.3 }.  NOTE: Since the termini are dependent on
+  # the amino acid sequence, they are give the *differential* mass.  The
+  # termini are given the special symbol as in sequest e.g. '[' => 12.22, #
+  # cterminus    ']' => 14.55 # nterminus
+  attr_accessor :masses_by_diff_mod_hash
+  # a hash, key is [AA_one_letter_symbol.to_sym, difference.to_f]
+  # values are the special_symbols
+  attr_accessor :mod_symbols_hash
+
+  # The modification symbols string looks like this:
+  # (M* +15.90000) (M# +29.00000) (S@ +80.00000) (C^ +12.00000) (ct[ +12.33000) (nt] +14.20000)
+  # ct is cterminal peptide (differential)
+  # nt is nterminal peptide (differential)
+  # the C is just cysteine
+  # will set_modifications and masses_by_diff_mod hash
+  def initialize(params, modification_symbols_string='')
+    @params = params
+    set_modifications(params, modification_symbols_string)
+  end
+
+  # set the masses_by_diff_mod and mod_symbols_hash from 
+  def set_hashes(modification_symbols_string)
+    @mod_symbols_hash = {}
+    @masses_by_diff_mod = {}
+    if modification_symbols_string == nil || modification_symbols_string == ''
+      return nil
+    end
+    table = @params.mass_table
+    modification_symbols_string.split(/\)\s+\(/).each do |mod|
+      if mod =~ /\(?(\w{1,2})(.) (.[\d\.]+)\)?/
+        aa_as_sym = $1.to_sym,
+        @mod_symbols_hash[[aa_as_sym, $3.to_f]] = $2.dup
+        if $1 == 'ct' || $1 == 'nt' 
+          @masses_by_diff_mod[$2] = $3.to_f
+        else
+          @masses_by_diff_mod[$1+$2] = $3.to_f + table[aa_as_sym]
+        end
+      end
+    end
+  end
+
+  # given a bare peptide (no end pieces) returns a ModificationInfo object
+  # e.g. given "]PEPT*IDE", NOT 'K.PEPTIDE.R'
+  # if there are no modifications, returns nil
+  def modification_info(peptide)
+    if @masses_by_diff_mod.size == 0
+      return nil
+    end
+    hash[:modified_peptide] = peptide.dup
+    hash = {}
+    hsh = @masses_by_diff_mod  
+    table = @params.mass_table
+    h = table[:h]  # this? or h_plus ??
+    oh = table[:o] + h
+    ## only the termini can match a single char
+    if hsh.key? peptide[0,1]
+      # AA + H + differential_mod
+      hash[:mod_nterm_mass] = table[peptide[1,1].to_sym] + h + hsh[peptide[0,1]]
+      peptide.slice!( 1..-1 )
+    end
+    if hsh.key? peptide[-1,1]
+      # AA + OH + differential_mod
+      hash[:mod_cterm_mass] = table[peptide[-2,1].to_sym] + oh + hsh[peptide[-1,1]]
+      peptide.slice!( 0..-2 )
+    end
+    mod_array = []
+    (0...peptide.size).each do |i|
+      if hsh.key? peptide[i,2]
+        mod_array << [ i+1 , hsh[peptide[i,2]] ]
+      end
+    end
+    if mod_array.size > 0
+      hash[:mod_aminoacid_mass_array] = mod_array
+    end
+    if hash.size > 0
+      SpecID::Sequest::PepXML::SearchHit::ModificationInfo.new(hash)
+    else
+      nil
+    end
+  end
+
+  # 1. sets aa_mods and term_mods from a sequest params object
+  # 2. sets @params
+  # 3. sets @masses_by_diff_mod
+  def set_modifications(params, modification_symbols_string)
+    @params = params
+
+    set_hashes(modification_symbols_string)
+
+    ####################################
+    ## static mods
+    ####################################
+
+    static_mods = [] # [[one_letter_amino_acid.to_sym, add_amount.to_f], ...]
+    static_terminal_mods = [] # e.g. [add_Cterm_peptide, amount.to_f]
+
+    params.mods.each do |k,v|
+      v_to_f = v.to_f
+      if v_to_f != 0.0
+        if k =~ /add_(\w)_/
+          static_mods << [$1.to_sym, v_to_f]
+        else
+          static_terminal_mods << [k, v_to_f]
+        end
+      end
+    end
+    aa_hash = params.mass_table
+
+    ## Create the static_mods objects
+    static_mods.map! do |mod|
+      hash = {
+        :aminoacid => mod[0].to_s,
+        :massdiff => mod[1].to_plus_minus_string,
+        :mass => aa_hash[mod[0]] + mod[1],
+        :variable => 'N',
+        :binary => 'Y',
+      } 
+      SpecID::Sequest::PepXML::AAModification.new(hash)
+    end
+
+    ## Create the static_terminal_mods objects
+    static_terminal_mods.map! do |mod|
+      terminus = if mod[0] =~ /Cterm/ ; 'c'
+                 else                 ; 'n' # only two possible termini
+                 end
+      protein_terminus = case mod[0] 
+                         when /Nterm_protein/ ; 'n'
+                         when /Cterm_protein/ ; 'c'
+                         else nil
+                         end
+
+      # create the hash                            
+      hash = {
+        :terminus => terminus,
+        :massdiff => mod[1].to_plus_minus_string,
+        :variable => 'N',
+        :description => mod[0],
+      }
+      hash[:protein_terminus] = protein_terminus if protein_terminus
+      SpecID::Sequest::PepXML::TerminalModification.new(hash)
+    end
+    #################################
+    # Variable Mods:
+    #################################
+    arr = params.diff_search_options.rstrip.split(/\s+/)
+    # [aa.to_sym, diff.to_f]
+    variable_mods = []
+    (0...arr.size).step(2) do |i|
+      if arr[i].to_f != 0.0
+        variable_mods << [arr[i+1].to_sym, arr[i].to_f]
+      end
+    end
+    variable_mods.map! do |mod|
+      hash = {
+        :aminoacid => mod[0].to_s,
+        :massdiff => mod[1].to_plus_minus_string,
+        :mass => aa_hash[mod[0]] + mod[1],
+        :variable => 'Y',
+        :binary => 'N',
+        :symbol => @mod_symbols_hash[mod],
+      }
+      SpecID::Sequest::PepXML::AAModification.new(hash)
+    end
+    #################################
+    # TERMINAL Variable Mods:
+    #################################
+    # These are always peptide, not protein termini (for sequest)
+    (nterm_diff, cterm_diff) = params.term_diff_search_options.rstrip.split(/\s+/).map{|v| v.to_f }
+
+    to_add = []
+    if nterm_diff != 0.0
+      to_add << ['n',nterm_diff.to_plus_minus_string, @mod_symbols_hash[:nt, nterm_diff]]
+    end
+    if cterm_diff != 0.0
+      to_add << ['c', cterm_diff.to_plus_minus_string, @mod_symbols_hash[:ct, cterm_diff]]
+    end
+
+    variable_terminal_mods = to_add.map do |term, mssdiff, symb|
+      hash = {
+        :terminus => term,
+        :massdiff => mssdiff,
+        :variable => 'Y',
+        :symbol => symb,
+      }
+      SpecID::Sequest::PepXML::TerminalModification.new(hash)
+    end
+
+    #########################
+    # COLLECT THEM
+    #########################
+    @aa_mods = static_mods + variable_mods
+    @term_mods = static_terminal_mods + variable_terminal_mods
+  end
+
+  ## Generates the pepxml for static and differential amino acid mods based on
+  ## sequest object
+  def to_pepxml
+    st = ''
+    if @aa_mods
+      st << @aa_mods.map {|v| v.to_pepxml }.join
+    end
+    if @term_mods
+      st << @term_mods.map {|v| v.to_pepxml }.join
+    end
+    st
+  end
+
+end
+
+# Modified aminoacid, static or variable
+# unless otherwise stated, all attributes can be anything
+class SpecID::Sequest::PepXML::AAModification
+  include SpecIDXML
+
+  # The amino acid (one letter code)
+  attr_accessor :aminoacid
+  # Must be a string!!!!
+  # Mass difference with respect to unmodified aminoacid, must begin with
+  # either + (nonnegative) or - [e.g. +1.05446 or -2.3342]
+  # consider Numeric#to_plus_minus_string at top
+  attr_accessor :massdiff
+  # Mass of modified aminoacid
+  attr_accessor :mass
+  # Y if both modified and unmodified aminoacid could be present in the
+  # dataset, N if only modified aminoacid can be present
+  attr_accessor :variable
+  # whether modification can reside only at protein terminus (specified 'n',
+  # 'c', or 'nc')
+  attr_accessor :peptide_terminus
+  # Special symbol used by search engine to designate this modification
+  attr_accessor :symbol
+  # Y if each peptide must have only modified or unmodified aminoacid, N if a
+  # peptide may contain both modified and unmodified aminoacid
+  attr_accessor :binary
+
+  def initialize(hash=nil)
+    instance_var_set_from_hash(hash) if hash # can use unless there are weird methods
+  end
+
+  def to_pepxml
+    short_element_xml_from_instance_vars("aminoacid_modification")
+  end
+
+end
+
+# Modified aminoacid, static or variable
+class SpecID::Sequest::PepXML::TerminalModification
+  include SpecIDXML
+
+  # n for N-terminus, c for C-terminus
+  attr_accessor :terminus
+  # Mass difference with respect to unmodified terminus
+  attr_accessor :massdiff
+  # Mass of modified terminus
+  attr_accessor :mass
+  # Y if both modified and unmodified terminus could be present in the
+  # dataset, N if only modified terminus can be present
+  attr_accessor :variable
+  # Special symbol used by search engine to designate this modification
+  attr_accessor :symbol
+  # whether modification can reside only at protein terminus (specified n or
+  # c)
+  attr_accessor :protein_terminus
+  attr_accessor :description
+
+  def initialize(hash=nil)
+    instance_var_set_from_hash(hash) if hash # can use unless there are weird methods
+  end
+
+  def to_pepxml
+    short_element_xml_from_instance_vars("terminal_modification")
+  end
+end
+
+
 class SpecID::Sequest::PepXML::SearchDatabase
   include SpecIDXML 
   attr_accessor :local_path
@@ -708,7 +1318,15 @@ end
 
 class SpecID::Sequest::PepXML::SpectrumQuery
   include SpecIDXML
-  attr_accessor :spectrum, :start_scan, :end_scan, :precursor_neutral_mass, :index, :search_results
+
+  # basename_noext.first_scan.last_scan.charge
+  attr_accessor :spectrum
+  attr_accessor :start_scan
+  attr_accessor :end_scan
+  attr_accessor :precursor_neutral_mass
+  attr_accessor :index
+  attr_accessor :search_results
+
   # this is a string
   attr_accessor :assumed_charge
   attr_accessor :pepxml_version
@@ -803,6 +1421,10 @@ end
 # this responds to flatten (so that it won't flatten).
 class SpecID::Sequest::PepXML::SearchHit < Array
   include SpecIDXML
+
+  Non_standard_amino_acid_char_re = /[^A-Z\.\-]/
+
+  # num_tot_proteins = "Number of unique proteins in search database containing peptide"
   #attr_accessor 0:hit_rank, 1:peptide, 2:peptide_prev_aa, 3:peptide_next_aa, 4:protein, 5:num_tot_proteins, 6:num_matched_ions, 7:tot_num_ions, 8:calc_neutral_pep_mass, 9:massdiff, 10:num_tol_term, 11:num_missed_cleavages, 12:is_rejected
   #attr_accessor 13:deltacnstar
   #attr_accessor 14:xcorr, 15:deltacn, 16:spscore, 17:sprank
@@ -811,69 +1433,82 @@ class SpecID::Sequest::PepXML::SearchHit < Array
   ind_keys = {:hit_rank => 0, :peptide => 1, :peptide_prev_aa => 2, :peptide_next_aa => 3, :protein => 4, :num_tot_proteins => 5, :num_matched_ions => 6, :tot_num_ions => 7, :calc_neutral_pep_mass => 8, :massdiff => 9, :num_tol_term => 10, :num_missed_cleavages => 11, :is_rejected => 12, :deltacnstar  => 13, :xcorr => 14, :deltacn => 15, :spscore => 16, :sprank => 17}
   @@methods = ind_keys.keys
   def hit_rank ; self[0] end ; def hit_rank=(oth) ; self[0] = oth end
-def peptide ; self[1] end ; def peptide=(oth) ; self[1] = oth end
-def peptide_prev_aa ; self[2] end ; def peptide_prev_aa=(oth) ; self[2] = oth end
-def peptide_next_aa ; self[3] end ; def peptide_next_aa=(oth) ; self[3] = oth end
-def protein ; self[4] end ; def protein=(oth) ; self[4] = oth end
-def num_tot_proteins ; self[5] end ; def num_tot_proteins=(oth) ; self[5] = oth end
-def num_matched_ions ; self[6] end ; def num_matched_ions=(oth) ; self[6] = oth end
-def tot_num_ions ; self[7] end ; def tot_num_ions=(oth) ; self[7] = oth end
-def calc_neutral_pep_mass ; self[8] end ; def calc_neutral_pep_mass=(oth) ; self[8] = oth end
-def massdiff ; self[9] end ; def massdiff=(oth) ; self[9] = oth end
-def num_tol_term ; self[10] end ; def num_tol_term=(oth) ; self[10] = oth end
-def num_missed_cleavages ; self[11] end ; def num_missed_cleavages=(oth) ; self[11] = oth end
-def is_rejected ; self[12] end ; def is_rejected=(oth) ; self[12] = oth end
-def deltacnstar ; self[13] end ; def deltacnstar=(oth) ; self[13] = oth end
-def xcorr ; self[14] end ; def xcorr=(oth) ; self[14] = oth end
-def deltacn ; self[15] end ; def deltacn=(oth) ; self[15] = oth end
-def spscore ; self[16] end ; def spscore=(oth) ; self[16] = oth end
-def sprank ; self[17] end ; def sprank=(oth) ; self[17] = oth end
-
-@@arr_size = ind_keys.size
-ind_keys.each {|k,v| ind_keys_w_eq["#{k}=".to_sym] = v }
-ind_keys.merge!(ind_keys_w_eq)
-ind_keys.each {|k,v| @@ind[k] = v ; @@ind["#{k}"] = v}
-
-# These are all search_score elements:
-
-# 1 if there is no second ranked hit, 0 otherwise
-
-def initialize(hash=nil)
-  super(@@arr_size)
-  self[0,18] = [hash[:hit_rank], hash[:peptide], hash[:peptide_prev_aa], hash[:peptide_next_aa], hash[:protein], hash[:num_tot_proteins], hash[:num_matched_ions], hash[:tot_num_ions], hash[:calc_neutral_pep_mass], hash[:massdiff], hash[:num_tol_term], hash[:num_missed_cleavages], hash[:is_rejected], hash[:deltacnstar], hash[:xcorr], hash[:deltacn], hash[:spscore], hash[:sprank]]
-  self
-  #if hash ; set_from_hash(hash) end
-end
+  def peptide ; self[1] end ; def peptide=(oth) ; self[1] = oth end
+  def peptide_prev_aa ; self[2] end ; def peptide_prev_aa=(oth) ; self[2] = oth end
+  def peptide_next_aa ; self[3] end ; def peptide_next_aa=(oth) ; self[3] = oth end
+  def protein ; self[4] end ; def protein=(oth) ; self[4] = oth end
+  def num_tot_proteins ; self[5] end ; def num_tot_proteins=(oth) ; self[5] = oth end
+  def num_matched_ions ; self[6] end ; def num_matched_ions=(oth) ; self[6] = oth end
+  def tot_num_ions ; self[7] end ; def tot_num_ions=(oth) ; self[7] = oth end
+  def calc_neutral_pep_mass ; self[8] end ; def calc_neutral_pep_mass=(oth) ; self[8] = oth end
+  def massdiff ; self[9] end ; def massdiff=(oth) ; self[9] = oth end
+  def num_tol_term ; self[10] end ; def num_tol_term=(oth) ; self[10] = oth end
+  def num_missed_cleavages ; self[11] end ; def num_missed_cleavages=(oth) ; self[11] = oth end
+  def is_rejected ; self[12] end ; def is_rejected=(oth) ; self[12] = oth end
+  def deltacnstar ; self[13] end ; def deltacnstar=(oth) ; self[13] = oth end
+  def xcorr ; self[14] end ; def xcorr=(oth) ; self[14] = oth end
+  def deltacn ; self[15] end ; def deltacn=(oth) ; self[15] = oth end
+  def spscore ; self[16] end ; def spscore=(oth) ; self[16] = oth end
+  def sprank ; self[17] end ; def sprank=(oth) ; self[17] = oth end
+
+  @@arr_size = ind_keys.size
+  ind_keys.each {|k,v| ind_keys_w_eq["#{k}=".to_sym] = v }
+  ind_keys.merge!(ind_keys_w_eq)
+  ind_keys.each {|k,v| @@ind[k] = v ; @@ind["#{k}"] = v}
+
+  # These are all search_score elements:
+
+  # 1 if there is no second ranked hit, 0 otherwise
 
-# Returns prev, peptide, next from sequence.  Parse errors return
-# nil,nil,nil
-#   R.PEPTIDE.A  # -> R, PEPTIDE, A
-#   R.PEPTIDE.-  # -> R, PEPTIDE, -
-#   PEPTIDE.A    # -> -, PEPTIDE, A
-#   A.PEPTIDE    # -> A, PEPTIDE, -
-#   PEPTIDE      # -> nil,nil,nil
-def self.split_sequence(val)
-  peptide_prev_aa = ""; peptide = ""; peptide_next_aa = ""
-  pieces = val.split(".") 
-  case pieces.size
-  when 3
-    peptide_prev_aa, peptide, peptide_next_aa = *pieces
-  when 2
-    if pieces[0].size > 1  ## N termini
-      peptide_prev_aa, peptide, peptide_next_aa = '-', pieces[0], pieces[1]
-    else  ## C termini
-      peptide_prev_aa, peptide, peptide_next_aa = pieces[0], pieces[1], '-'
-    end
-  when 1  ## this must be a parse error!
-    peptide_prev_aa, peptide, peptide_next_aa = nil,nil,nil
-  when 0
-    peptide_prev_aa, peptide, peptide_next_aa = nil,nil,nil
-  end
-  return peptide_prev_aa, peptide, peptide_next_aa
-end
+  def initialize(hash=nil)
+    super(@@arr_size)
+    self[0,18] = [hash[:hit_rank], hash[:peptide], hash[:peptide_prev_aa], hash[:peptide_next_aa], hash[:protein], hash[:num_tot_proteins], hash[:num_matched_ions], hash[:tot_num_ions], hash[:calc_neutral_pep_mass], hash[:massdiff], hash[:num_tol_term], hash[:num_missed_cleavages], hash[:is_rejected], hash[:deltacnstar], hash[:xcorr], hash[:deltacn], hash[:spscore], hash[:sprank]]
+    self
+    #if hash ; set_from_hash(hash) end
+  end
+
+  # remove_non_amino_acids && split_sequence
+  def self.prepare_sequence(val)
+    nv = remove_non_amino_acids(val)
+    split_sequence(nv)
+  end
 
-def inspect
-    "#<SearchHit #{@@methods.map do |m| "#{m}:#{self.send(m)}" end.join(" ")}>"
+  # Returns prev, peptide, next from sequence.  Parse errors return
+  # nil,nil,nil
+  #   R.PEPTIDE.A  # -> R, PEPTIDE, A
+  #   R.PEPTIDE.-  # -> R, PEPTIDE, -
+  #   PEPTIDE.A    # -> -, PEPTIDE, A
+  #   A.PEPTIDE    # -> A, PEPTIDE, -
+  #   PEPTIDE      # -> nil,nil,nil
+  def self.split_sequence(val)
+    peptide_prev_aa = ""; peptide = ""; peptide_next_aa = ""
+    pieces = val.split('.') 
+    case pieces.size
+    when 3
+      peptide_prev_aa, peptide, peptide_next_aa = *pieces
+    when 2
+      if pieces[0].size > 1  ## N termini
+        peptide_prev_aa, peptide, peptide_next_aa = '-', pieces[0], pieces[1]
+      else  ## C termini
+        peptide_prev_aa, peptide, peptide_next_aa = pieces[0], pieces[1], '-'
+      end
+    when 1  ## this must be a parse error!
+      peptide_prev_aa, peptide, peptide_next_aa = nil,nil,nil
+    when 0
+      peptide_prev_aa, peptide, peptide_next_aa = nil,nil,nil
+    end
+    return peptide_prev_aa, peptide, peptide_next_aa
+  end
+
+  # removes nonstandard chars with Non_standard_amino_acid_char_re
+  # preserves A-Z and '.
+  def self.remove_non_amino_acids(sequence)
+    sequence.gsub(Non_standard_amino_acid_char_re, '')
+  end
+
+  def inspect
+    var = @@methods.map do |m| "#{m}:#{self.send(m)}" end.join(" ")
+    "#<SearchHit #{var}>"
   end
 
   # requires Params object and full sequence (with heads and tails)
@@ -924,3 +1559,65 @@ def inspect
 
 end
 
+# Positions and masses of modifications
+class SpecID::Sequest::PepXML::SearchHit::ModificationInfo
+  include SpecIDXML
+
+  ## Should be something like this:
+  # <modification_info mod_nterm_mass=" " mod_nterm_mass=" " modified_peptide=" ">
+  #   <mod_aminoacid_mass position=" " mass=" "/>
+  # </modification_info>
+
+
+  # Mass of modified N terminus<
+  attr_accessor :mod_nterm_mass
+  # Mass of modified C terminus<
+  attr_accessor :mod_cterm_mass
+  # Peptide sequence (with indicated modifications)  I'm assuming that the
+  # native sequest indicators are OK here
+  attr_accessor :modified_peptide
+  ## A few main types:
+
+  # this should be an array of arrays: [[position, modified_mass], ...]
+  # position ranges from 1 to peptide length
+  attr_accessor :mod_aminoacid_mass_array
+
+  def initialize(hash=nil)
+    instance_var_set_from_hash(hash)
+  end
+
+  # Will escape any xml special chars in modified_peptide
+  def to_pepxml
+    ## Collect the modifications:
+    mod_strings = []
+    if @mod_aminoacid_mass_array
+      mod_strings = @mod_aminoacid_mass_array.map do |ar|
+        "position=\"#{ar[0]}\" mass=\"#{ar[1]}\""
+      end
+    end
+    ## Create the attribute string:
+    att_parts = []
+    if @mod_nterm_mass
+      att_parts << "mod_nterm_mass=\"#{@mod_nterm_mass}\""
+    end
+    if @mod_cterm_mass
+      att_parts << "mod_cterm_mass=\"#{@mod_cterm_mass}\""
+    end
+    if @modified_peptide
+      att_parts << "modified_peptide=\"#{escape_special_chars(@modified_peptide)}\""
+    end
+    element_xml_and_att_string('modification_info', att_parts.join(" ")) do
+      mod_strings.map {|st| short_element_xml_and_att_string('mod_aminoacid_mass', st) }.join
+    end
+  end
+
+  ## 
+
+  # <modification_info modified_peptide="GC[546]M[147]PSKEVLSAGAHR">
+  # <mod_aminoacid_mass position="2" mass="545.7160"/>
+  # <mod_aminoacid_mass position="3" mass="147.1926"/>
+  # </modification_info>
+
+  
+end
+