mspire 0.3.1 → 0.3.9

data/lib/validator/decoy.rb

@@ -3,10 +3,12 @@ require 'validator'
 class Validator::Decoy < Validator
   include Precision::Calculator::Decoy
 
+  # a Regexp (if concatenated) or a String (the filename of separate run)
   attr_accessor :constraint
 
   attr_accessor :decoy_on_match
   attr_accessor :correct_wins
+  attr_accessor :decoy_to_target_ratio
 
   attr_accessor :last_pep_was_decoy
 
@@ -16,13 +18,19 @@ class Validator::Decoy < Validator
 
   attr_reader :normal_peps_just_submitted
 
-  def initialize(constraint=nil, decoy_on_match = true, correct_wins = true)
-    @decoy_on_match = decoy_on_match
-    @correct_wins = correct_wins
-    @constraint = constraint
+  DEFAULTS = {
+    :decoy_on_match => true,
+    :correct_wins => true,
+    :decoy_to_target_ratio => 1.0,
+  }
+
+  def initialize(opts={})
+    merged = DEFAULTS.merge(opts)
+    @constraint, @decoy_on_match, @correct_wins, @decoy_to_target_ratio = merged.values_at(:constraint, :decoy_on_match, :correct_wins, :decoy_to_target_ratio)
   end
 
   # returns [normal, decoy] (?? I think ??)
+  # reads the full protein reference
   def partition(peps)
     if @decoy_on_match 
       if @correct_wins
@@ -74,15 +82,15 @@ class Validator::Decoy < Validator
     @normal_peps_just_submitted = normal
     @increment_normal += normal.size
     @increment_decoy += decoy.size
-    calc_precision(@increment_normal, @increment_decoy)
+    calc_precision(@increment_normal, @increment_decoy, @decoy_to_target_ratio)
   end
 
   def pephit_precision(peps, separate_peps=nil)
     if separate_peps
-      calc_precision(peps.size, separate_peps.size)
+      calc_precision(peps.size, separate_peps.size, @decoy_to_target_ratio)
     else
       (norm, decoy) = partition(peps)
-      calc_precision(norm.size, decoy.size)
+      calc_precision(norm.size, decoy.size, @decoy_to_target_ratio)
     end
   end
 

data/lib/validator/digestion_based.rb

@@ -6,7 +6,8 @@ require 'spec_id/sequest/params'
 # SpecID::Pep objects using the pephit_precision method.
 class Validator::DigestionBased < Validator
   DEFAULTS = {
-    :false_to_total_ratio => 1.0,
+    #:false_to_total_ratio => 1.0,  # disable because this needs to be set
+    # explicitly
     :background => 0.0,
   }
 
@@ -42,13 +43,13 @@ class Validator::DigestionBased < Validator
   # returns [num_tps, num_fps]
   def calc_precision_prep(num_tps, num_fps)
     total_peps_passing_partition = num_tps + num_fps
-    num_fps = adjust_fps_for_background(num_tps, num_fps, @background)
+    num_fps = adjust_fps_for_background(num_tps, num_fps, background)
     ## we must use the false_to_total_ratio to estimate how many are really
     ## incorrect!
     # FALSE/TOTAL  = FALSE(found)/TOTAL(found)
     # TOTAL(found) = FALSE(found) * TOTAL/FALSE
     #              = FALSE(found) / (FALSE/TOTAL)
-    total_false = num_fps / @false_to_total_ratio
+    total_false = num_fps / false_to_total_ratio
     # NOTE: the partition algorithm drops peptides that are transmembrane
     # under certain options.  Thus, the total false estimate must be tempered
     # by this lower number of total peptides.
@@ -60,7 +61,7 @@ class Validator::DigestionBased < Validator
   # assumes partition returns (tps, fps)
   def set_false_to_total_ratio(peps)
     (tps, fps) = partition(peps)
-    @false_to_total_ratio = fps.size.to_f / (tps.size + fps.size) 
+    self.false_to_total_ratio = fps.size.to_f / (tps.size + fps.size) 
     self
   end
 

data/lib/validator/q_value.rb

@@ -0,0 +1,32 @@
+
+
+# from percolator
+# This is a trivial class (since q-values are so straightforward with regards
+# to precision), but it allows us to work with q-values using the same
+# interface as all other validators
+class Validator::QValue
+
+  # objs should respond_to :q_value
+  # q-values: 0.0 means no false discoveries, 0.5 means 50% false discoveries
+  # 1 - (the largest q value) is the precision
+  def precision(objs)
+    return 1.0 if objs.size == 0
+    largest_q_value = objs.map {|v| v.q_value }.max
+    prec = 1.0 - largest_q_value
+  end
+
+
+  # objs should respond_to :q_value
+  # These should be added from low q-value to high q-value
+  # The last q-value added determines the precision
+  def increment_precision(objs)
+    if objs.is_a?(SpecID::Pep) or objs.is_a?(SpecID::Prot)
+      objs = [objs]
+    end
+    precision(objs)
+  end
+
+  alias_method :pephit_precision, :precision
+  alias_method :prothit_precision, :precision
+  alias_method :increment_pephits_precision, :increment_precision
+end

data/script/peps_per_bin.rb

@@ -0,0 +1,67 @@
+#!/usr/bin/ruby -w
+
+require 'generator'
+require 'optparse'
+
+require 'fasta'
+require 'sample_enzyme'
+require 'spec_id/digestor'
+require 'spec_id/mass'
+require 'vec'
+
+opt = {}
+opt[:missed_cleavages] = 0 # ~ parts per million
+opt[:bin_size] = 0.001  # ~ parts per million
+opt[:min] = 300.0
+opt[:max] = 4500.0
+opt[:h_plus] = 1.0
+
+opts = OptionParser.new do |op|
+  op.banner = "usage: #{File.basename(__FILE__)} *.fasta"
+  op.separator "Outputs a close estimate of number of peptides per bin."
+  op.separator "Uses m+H+ as the peptide mass."
+  op.separator "[for speed, assumes that there is a peptide mass close to the extremes]"
+  op.on("-b", "--bin_size <F>", Float, "size of bins [#{opt[:bin_size]}]") {|v| opt[:bin_size] = v }
+  op.on("-x", "--max <F>", Float, "max mass to accept [#{opt[:max]}]") {|v| opt[:max] = v }
+  op.on("-n", "--min <F>", Float, "min mass to accept [#{opt[:min]}]") {|v| opt[:min] = v }
+  op.on("-h", "--h_plus <F>", Float, "value of H+ to use [#{opt[:h_plus]}]") {|v| opt[:h_plus] = v }
+  op.on("-m", "--missed_cleavages <N>", Integer, "num missed cleavages [#{opt[:missed_cleavages]}]") {|v| opt[:missed_cleavages] = v }
+end
+
+opts.parse!
+
+if ARGV.size == 0
+  puts opts.to_s
+  exit
+end
+
+min_mass = opt[:min]
+max_mass = opt[:max]
+
+ARGV.each do |file|
+  fasta = Fasta.new(file)
+  uniq_aaseqs = fasta.map do |prot|
+    SampleEnzyme.tryptic(prot.aaseq, opt[:missed_cleavages])
+  end.flatten.uniq
+
+  masses = Mass::Calculator.new(Mass::MONO, opt[:h_plus]).masses(uniq_aaseqs)
+  passing_masses = Mass::Calculator.new(Mass::MONO, opt[:h_plus]).masses(uniq_aaseqs).select do |mh|
+    ((mh >= min_mass) and (mh <= max_mass))
+  end
+
+  ## warn if the masses aren't close to the end points
+  if (max_mass - passing_masses.max) > 1.0
+    warn "highest mass is not that close to max: #{passing_masses.max}"
+  end
+  if (passing_masses.min - min_mass) > 1.0
+    warn "lowest mass is not that close to min: #{passing_masses.min}"
+  end
+
+  num_bins = (max_mass - min_mass) / opt[:bin_size]
+
+  (bins, freqs) = VecD.new(passing_masses).histogram(num_bins)
+
+  # report
+  puts "#{file}: #{freqs.avg}"
+
+end

data/script/sqt_to_meta.rb

@@ -0,0 +1,24 @@
+#!/usr/bin/ruby -s
+
+require 'optparse'
+
+$outfile = 'meta.sqm'
+opts = OptionParser.new do |op|
+  op.banner = "usage: #{File.basename(__FILE__)} <file>.sqt ..."
+  op.separator "outputs meta.sqm (a sqt meta file)"
+  op.on("-o", "--outfile <file>", "currently: #{$outfile}") {|v| $outfile = v}
+end
+
+opts.parse!
+
+if ARGV.size == 0
+  puts opts.to_s
+  exit
+end
+
+File.open($outfile, 'w') do |out|
+  ARGV.each do |file|
+    out.puts File.expand_path(file)
+  end
+end
+

data/specs/bin/bioworks_to_pepxml_spec.rb

@@ -41,7 +41,7 @@ describe 'bioworks_to_pepxml.rb' do
       cmd = "#{@cmd} -p #{@tf_params} -o #{@out_path} #{@tf_bioworks_xml} -m #{@tf_mzxml_path} -d /work/special/path --copy_mzxml"
       ## FILES EXIST:
       prc = proc {|file| 
-        file.should exist
+        file.exist_as_a_file?.should be_true
         beginning = IO.readlines(file)[0,50].join("\n")
         $XML_SANITY_LINES.each do |line|
           beginning.should include(line)
@@ -55,7 +55,7 @@ describe 'bioworks_to_pepxml.rb' do
       ## COPY MZXML:
       %w(000 020).each do |file|
         mzxml_file = File.join(@out_path, "#{file}.mzXML")
-        mzxml_file.should exist
+        mzxml_file.exist_as_a_file?.should be_true
       end
       ## CLEANUP:
       unless @no_delete then FileUtils.rm_rf(@out_path) end
@@ -68,7 +68,7 @@ describe 'bioworks_to_pepxml.rb' do
       db_re = /C:\\Xcalibur\\database\\ecoli_K12_ncbi_20060321.fasta/
       IO.read(@tf_params).should =~ db_re
       prc = proc {|file|
-        file.should exist
+        file.exist_as_a_file?.should be_true
         IO.read(file).should_not =~ db_re
       }
       _basic(cmd, prc)

data/specs/bin/fasta_shaker_spec.rb

@@ -200,13 +200,13 @@ EDITPEP
     end
 
   def fastalns(fn)
-    fn.should exist
+    fn.exist_as_a_file?.should be_true
     IO.read(fn).split("\n")
   end
 
   # returns the fasta object proteins
   def fastap(fn)
-    @f.should exist
+    @f.exist_as_a_file?.should be_true
     Fasta.new(fn).prots
   end
 

data/specs/bin/filter_and_validate__multiple_vals_helper.yaml

@@ -2,10 +2,10 @@
 pephits_precision: 
 - validator: decoy
   value: 0.992932862190813
-- validator: badAA
+- validator: badAAEst
   value: 0.178006237270664
-- validator: badAA
-  value: -0.0247654296463377
+- validator: badAAEst
+  value: -0.0247654296463379 
 - validator: badAA
   value: 0.301413862599215
 - validator: bias
@@ -94,22 +94,19 @@ params:
     :decoy_on_match: true
     :correct_wins: true
   - :calculated_background: 0.127208480565371
-    :type: badAA
-    :class: Validator::AA
+    :type: badAAEst
+    :class: Validator::AAEst
     :background: 0.001
     :frequency: 0.0147528119278054
-    :false_to_total_ratio: 1.0
   - :calculated_background: 0.402826855123675
-    :type: badAA
-    :class: Validator::AA
+    :type: badAAEst
+    :class: Validator::AAEst
     :background: 0.0
     :frequency: 0.0463510332199843
-    :false_to_total_ratio: 1.0
   - :calculated_background: 0.127208480565371
     :type: badAA
     :class: Validator::AA
     :background: 0.001
-    :frequency: 
     :false_to_total_ratio: 0.180662732637313
   - :calculated_background: 0.773851590106007
     :type: bias

data/specs/bin/filter_and_validate_spec.rb

@@ -1,3 +1,5 @@
+require 'yaml'
+
 require File.expand_path( File.dirname(__FILE__) + '/../spec_helper' )
 
 require 'spec_id/precision/filter'
@@ -80,7 +82,7 @@ describe 'filter_and_validate.rb on small bioworks file' do
       `#{run_normal}`
     end
     structs = [ht_file, hs_file].map do |file| 
-      file.should exist
+      file.exist_as_a_file?.should be_true
       struct = YAML.load_file(file)
       File.unlink file
       struct
@@ -104,7 +106,7 @@ describe 'filter_and_validate.rb on small bioworks file' do
 
   it 'handles multiple validators of the same kind (except, of course, decoy)' do
 
-    struct = @st_to_yaml.call( "#{@fake_bioworks_file} --proteins  -1 0.0 -2 0.0 -3 0.0 -d 0.01 -p 1000000 --decoy /^DECOY_/ --digestion #{@small_fasta_file},#{@params_file} --bad_aa C,true,0.001 --bad_aa E,true --bad_aa C,false,0.001 --bias #{@small_bias_fasta_file},true --bias #{@small_bias_fasta_file},false --bias #{@small_bias_fasta_file},true,0.2 --tmm #{@phobius_file},1,true,0.8,0.2 --tmm #{@phobius_file} --tmm #{@toppred_file},3,true,false --tmm #{@toppred_file} --tps #{@small_bias_fasta_file} -o text_table:#{@table_output_file} " )
+    struct = @st_to_yaml.call( "#{@fake_bioworks_file} --proteins  -1 0.0 -2 0.0 -3 0.0 -d 0.01 -p 1000000 --decoy /^DECOY_/ --digestion #{@small_fasta_file},#{@params_file} --bad_aa_est C,0.001 --bad_aa_est E --bad_aa C,0.001 --bias #{@small_bias_fasta_file},true --bias #{@small_bias_fasta_file},false --bias #{@small_bias_fasta_file},true,0.2 --fasta #{@small_fasta_file} --tmm #{@phobius_file},1,true,0.8,0.2 --tmm #{@phobius_file} --tmm #{@toppred_file},3,true,false --tmm #{@toppred_file} --tps #{@small_bias_fasta_file} -o text_table:#{@table_output_file} " )
     frozen = YAML.load_file( File.dirname(__FILE__) + "/filter_and_validate__multiple_vals_helper.yaml" )
 
     ## Pephits precision:
@@ -121,8 +123,25 @@ describe 'filter_and_validate.rb on small bioworks file' do
     frp = frozen['params']
     stp = struct['params']
 
+    #puts "frozen validators:"
+    #p frp['validators']
+
+    #puts "seen validators:"
+    #p stp['validators']
+
     frp['validators'].zip(stp['validators']) do |f,s|
-      f.should == s
+      if f.is_a? Hash
+        f.keys.each do |k|
+          if k == :file or k == :transmem_file
+            File.basename(f[k]).should == File.basename(s[k].gsub('\\','/'))
+          else
+            s[k].should == f[k]
+            #f[k].should == s[k]
+          end
+        end
+      else
+        f.should == s
+      end
     end
 
     %w(ties prefilter top_hit_by decoy_on_match postfilter include_ties_in_top_hit_postfilter hits_together proteins include_ties_in_top_hit_prefilter).each do |k|
@@ -148,9 +167,9 @@ describe 'filter_and_validate.rb on small bioworks file' do
     text_table = IO.read(@table_output_file)
 
     # frozen
-    headings_re = Regexp.new( %w(num decoy badAA badAA badAA  bias  bias  bias   tmm   tmm   tmm   tmm   tps).join("\\s+") )
-    data_re = Regexp.new( %w(peps 283 0.993 0.178 -0.025 0.301 0.195 -4.793 0.403 0.438 -0.267 -0.156 -0.020 0.226).join("\\s+") )
-    prot_re = Regexp.new( %w(106 0.972 0.019   0.0 0.038 0.019   0.0 0.094 0.123   0.0   0.0   0.0 0.028).join("\\s+") )
+    headings_re = Regexp.new( %w(num decoy badAAEst badAAEst badAA  bias  bias  bias   tmm   tmm   tmm   tmm   tps).join("\\s+") )
+    data_re = Regexp.new( %w(peps 283 0.993 0.178006 -0.024765 0.301 0.195 -4.793 0.403 0.438 -0.267 -0.156 -0.020 0.226).join("\\s+") )
+    prot_re = Regexp.new( %w(106 0.972 0.018868   0.0 0.038 0.019   0.0 0.094 0.123   0.0   0.0   0.0 0.028).join("\\s+") )
     text_table.should =~ headings_re
     text_table.should =~ data_re
     text_table.should =~ prot_re

data/specs/bin/ms_to_lmat_spec.rb

@@ -16,7 +16,7 @@ describe 'ms_to_lmat.rb' do
     cmd = "#{@cmd} #{@mzxml} --ascii"   
     `#{cmd}`
     newfile = @mzxml.sub(".mzXML", ".lmata")
-    newfile.should exist
+    newfile.exist_as_a_file?.should be_true
     IO.read(newfile).should == IO.read(@ans_lmata)
     File.unlink(newfile)
   end
@@ -26,7 +26,7 @@ describe 'ms_to_lmat.rb' do
     cmd = "#{@cmd} #{@mzxml}"   
     `#{cmd}`
     newfile = @mzxml.sub(".mzXML", ".lmat")
-    newfile.should exist
+    newfile.exist_as_a_file?.should be_true
     IO.read(newfile).should == IO.read(@ans_lmat)
     File.unlink(newfile)
   end

data/specs/bin/prob_validate_spec.rb

@@ -1,3 +1,5 @@
+require 'yaml'
+
 require File.expand_path( File.dirname(__FILE__) + '/../spec_helper' )
 
 require 'spec_id/precision/prob'
@@ -47,19 +49,19 @@ describe 'filter_and_validate.rb on small bioworks file' do
   it 'responds to --prob init' do
     normal = @st_to_yaml.call( @args + " --prob" )
 
- normal[:pephits_precision].first[:values].zip([1.0, 1.0, 0.996655518394649, 0.918918918918919]) do |got,exp|
+ normal[:pephits_precision].first[:values].zip([1.0, 1.0, 0.993333333333333, 0.85]) do |got,exp|
       got.should be_close(exp, 0.000000000001)
     end
     #normal_nsp = @st_to_yaml.call( @args + " --prob nsp" )
     #normal.should == normal_nsp
     init = @st_to_yaml.call( @args + " --prob init" )
     init.should_not == normal
-    init[:pephits_precision].first[:values].zip([1.0, 0.974358974358974, 0.981324278438031, 0.890429958391123]) do |got,exp|
+    init[:pephits_precision].first[:values].zip([1.0, 0.95, 0.963333333333333, 0.8025]) do |got,exp|
       got.should be_close(exp, 0.000000000001)
     end
     with_sort_by = @st_to_yaml.call( @args + " --prob nsp --sort_by_init" )
     # frozen
-    with_sort_by[:pephits_precision].first[:values].zip([1.0, 0.994974874371859, 0.996655518394649, 0.918918918918919]) do |got,exp|
+    with_sort_by[:pephits_precision].first[:values].zip([1.0, 0.99, 0.993333333333333, 0.85]) do |got,exp|
       got.should be_close(exp, 0.000000000001)
     end
   end

data/specs/sample_enzyme_spec.rb

@@ -33,9 +33,94 @@ describe SampleEnzyme, "digesting sequences" do
     peps = SampleEnzyme.new('trypsin').digest(st, 2)
     peps.select {|aaseq| aaseq == 'CCCCK'}.size.should == 2
   end
-
   
 end
 
+describe SampleEnzyme, 'making enzyme calculations on sequences and aaseqs' do
+
+  before(:each) do
+    @full_KRP = SampleEnzyme.new do |se|
+      se.name = 'trypsin'
+      se.cut = 'KR'
+      se.no_cut = 'P'
+      se.sense = 'C'
+    end
+    @just_KR = SampleEnzyme.new do |se|
+      se.name = 'trypsin'
+      se.cut = 'KR'
+      se.no_cut = ''
+      se.sense = 'C'
+    end
+  end
+
+  it 'calculates the number of tolerant termini' do
+    exp = [{
+      # full KR/P
+      'K.EPTIDR.E' => 2,
+      'K.PEPTIDR.E' => 1,
+      'F.EEPTIDR.E' => 1,
+      'F.PEPTIDW.R' => 0,
+    },
+    {
+      # just KR
+      'K.EPTIDR.E' => 2,
+      'K.PEPTIDR.E' => 2,
+      'F.EEPTIDR.E' => 1,
+      'F.PEPTIDW.R' => 0,
+    }
+    ]
+    scall = Sequest::PepXML::SearchHit
+    sample_enzyme_ar = [@full_KRP, @just_KR]
+    sample_enzyme_ar.zip(exp) do |sample_enzyme,hash|
+      hash.each do |seq, val|
+        sample_enzyme.num_tol_term(seq).should == val
+      end
+    end
+  end
+
+  it 'calculates number of missed cleavages' do
+    exp = [{
+    "EPTIDR" => 0,
+    "PEPTIDR" => 0,
+    "EEPTIDR" => 0,
+    "PEPTIDW" => 0,
+    "PERPTIDW" => 0,
+    "PEPKPTIDW" => 0,
+    "PEPKTIDW" => 1,
+    "RTTIDR" => 1,
+    "RTTIKK" => 2,
+    "PKEPRTIDW" => 2,
+    "PKEPRTIDKP" => 2,
+    "PKEPRAALKPEERPTIDKW" => 3,
+    },
+    {
+    "EPTIDR" => 0,
+    "PEPTIDR" => 0,
+    "EEPTIDR" => 0,
+    "PEPTIDW" => 0,
+    "PERPTIDW" => 1,
+    "PEPKPTIDW" => 1,
+    "PEPKTIDW" => 1,
+    "RTTIDR" => 1,
+    "RTTIKK" => 2,
+    "PKEPRTIDW" => 2,
+    "PKEPRTIDKP" => 3,
+    "PKEPRAALKPEERPTIDKW" => 5,
+    }
+    ]
+
+    sample_enzyme_ar = [@full_KRP, @just_KR]
+    sample_enzyme_ar.zip(exp) do |sample_enzyme, hash|
+      hash.each do |aaseq, val|
+        #first, middle, last = SpecID::Pep.split_sequence(seq)
+        # note that we are only using the middle section!
+        sample_enzyme.num_missed_cleavages(aaseq).should == val
+      end
+    end
+  end
+
+end
+
+