RubyGems - mspire - Versions diffs - 0.1.5 → 0.1.7 - Mend

mspire 0.1.5 → 0.1.7

Files changed (47) hide show

data/Rakefile +5 -2
data/bin/bioworks_to_pepxml.rb +84 -40
data/bin/fasta_shaker.rb +100 -0
data/bin/filter_spec_id.rb +185 -23
data/bin/gi2annot.rb +2 -110
data/bin/id_class_anal.rb +31 -21
data/bin/id_precision.rb +12 -8
data/bin/{false_positive_rate.rb → precision.rb} +1 -1
data/bin/protein_summary.rb +55 -62
data/changelog.txt +34 -0
data/lib/align.rb +0 -1
data/lib/fasta.rb +88 -24
data/lib/gi.rb +114 -0
data/lib/roc.rb +64 -58
data/lib/spec_id/aa_freqs.rb +166 -0
data/lib/spec_id/bioworks.rb +5 -1
data/lib/spec_id/precision.rb +427 -0
data/lib/spec_id/proph.rb +2 -2
data/lib/spec_id/sequest.rb +810 -113
data/lib/spec_id/srf.rb +486 -0
data/lib/spec_id.rb +107 -23
data/release_notes.txt +11 -0
data/script/estimate_fpr_by_cysteine.rb +226 -0
data/script/filter-peps.rb +3 -3
data/script/find_cysteine_background.rb +137 -0
data/script/gen_database_searching.rb +11 -7
data/script/genuine_tps_and_probs.rb +136 -0
data/script/top_hit_per_scan.rb +5 -2
data/test/tc_aa_freqs.rb +59 -0
data/test/tc_bioworks.rb +6 -1
data/test/tc_bioworks_to_pepxml.rb +25 -18
data/test/tc_fasta.rb +81 -3
data/test/tc_fasta_shaker.rb +147 -0
data/test/tc_gi.rb +20 -0
data/test/tc_id_class_anal.rb +9 -12
data/test/tc_id_precision.rb +12 -11
data/test/{tc_false_positive_rate.rb → tc_precision.rb} +13 -22
data/test/tc_protein_summary.rb +31 -22
data/test/tc_roc.rb +95 -50
data/test/tc_sequest.rb +212 -145
data/test/tc_spec.rb +10 -5
data/test/tc_spec_id.rb +0 -2
data/test/tc_spec_id_xml.rb +36 -0
data/test/tc_srf.rb +216 -0
metadata +35 -21
data/lib/spec_id/false_positive_rate.rb +0 -476
data/test/tc_gi2annot.rb +0 -12

data/lib/spec_id/precision.rb ADDED Viewed

@@ -0,0 +1,427 @@
+require 'optparse'
+require 'ostruct'
+require 'generator'
+require 'gnuplot'
+require 'roc'
+class String
+  def margin
+    self.gsub(/^\s*\|/,'')
+  end
+end
+class SpecID ; end
+class SpecID::Precision ; end
+module SpecID::Precision::PlotHelper
+  PLOT_TYPE = 'XYData'
+  TITLE = 'Precision (Positive Predictive Value)'
+  XAXIS = 'Num Hits (excludes known false positives)'
+  EXT = '.toplot'
+  IMAGE_EXT = '.png'
+  def create_to_plot_file(all_arrs, key, files, filename_noext)
+    ## CREATE the PLOT IMAGE:
+    to_plot = filename_noext + EXT
+    png = filename_noext + IMAGE_EXT
+    File.open(to_plot,'w') do |out|
+      out.puts PLOT_TYPE
+      out.puts filename_noext
+      out.puts TITLE
+      out.puts XAXIS
+      out.puts escape_to_gnuplot(y_axis_label(key))
+      files.each_with_index do |file,i|
+        #p key[i]
+        #p all_arrs[i]
+        key[i].each_with_index do |k,j|
+          out.puts(escape_to_gnuplot("#{file}: #{k[1][1]}"))
+          out.puts all_arrs[i][j][0].join(' ')
+          out.puts all_arrs[i][j][1].join(' ')
+        end
+      end
+    end
+  end
+  ## outputs a .toplot file based on filename_noext, creates a png file, and
+  ## writes  html to fh that will load the png file up
+  ## This is a self contained module that can be swapped out for a
+  ## completely different plotting program if desired.
+  def plot_figure(all_arrs, key, files, filename_noext)
+    ## CREATE the PLOT IMAGE:
+    to_plot = filename_noext+'.toplot'
+    png = filename_noext+'.png'
+    Gnuplot.open do |gp|
+      Gnuplot::Plot.new( gp ) do |plot|
+        plot.terminal "png noenhanced"
+        plot.output png
+        plot.title TITLE
+        plot.xlabel XAXIS
+        plot.ylabel escape_to_gnuplot(y_axis_label(key))
+        plot.style "line 1 lt 1"
+        plot.style "line 2 lt 12"
+        #plot.style  "line 1 lt 1 lw #{opts.lw} pt 7 ps #{opts.ps}",
+        plot.yrange "[-0.05:#{1.05 + 0.020*files.size}]"
+        files.each_with_index do |file,i|
+          key[i].each_with_index do |k,j|
+            plot.data << Gnuplot::DataSet.new( [ all_arrs[i][j][0], all_arrs[i][j][1] ] ) do |ds|
+              ds.with = "lines"
+              ds.title = escape_to_gnuplot("#{file}: #{k[1][1]}")
+            end
+          end
+        end
+      end
+    end
+    ## CREATE the HTML to load the plot:
+    basename_filename_noext = File.basename(filename_noext)
+    output = "<div id=\"plot\"><table class=\"image\" align=\"center\">\n"
+    #output << "<caption align=\"bottom\">Additional views of this data may be obtained by using the <span class=\"code\">plot.rb</span> command on '#{to_plot}' (type <span class=\"code\">plot.rb</span> for more details). Plot generated with command: &nbsp;&nbsp; <span class=\"code\">#{plot_cmd}</span></caption>\n"
+    output << "<tr><td><img src=\"#{basename_filename_noext}.png\" title=\"File #{basename_filename_noext} must be in the same directory as this html.\"/></td></tr>\n"
+    output << "</table></div>\n"
+    output
+  end  # plot_figure
+end
+module SpecID::Precision::HTML
+  # html and body tags
+  def html
+        "|<html>
+         |#{yield}
+         |</html>\n".margin
+  end
+  def body
+        "|<body>
+         |  #{yield}
+         |</body>\n".margin
+  end
+  def header
+        "|<head>
+         |  #{style}
+         |</head>\n".margin
+  end
+  def td
+        "<td>#{yield}</td>"
+  end
+  def style
+        '
+     <style type="text/css">
+        div#tp_table {
+          text-align: center;
+          margin-top: 50px;
+          margin-bottom: 50px;
+        }
+        span.code {
+        font-family: Courier,Monospace;
+        font-size: 80%;
+        }
+          table {
+              border-width:1px;
+              border-color:#CCCCCC;
+              border-collapse: collapse;
+          }
+          caption {
+            font-size: 90%;
+          }
+          td,th {
+              padding-top: 2px;
+              padding-bottom: 2px;
+              padding-left: 1;
+              padding-right: 1;
+          }
+          th.small {
+            font-size: 80%;
+            font-weight: normal;
+            padding: 1px;
+          }
+          td.redline {
+              background-color: #FF0000;
+              color: #FFFFFF
+          }
+      div#plot {
+        margin: 30px;
+     text-align:center
+      }
+      hr {color: sienna}
+      body { font-size: 8pt; font-family: Arial,Helvetica,Times}
+      </style>
+        '
+  end
+  def table
+        "|<table border=\"1\" align=\"center\" style=\"font-size:100%\">
+         |  #{yield}
+         |</table>\n".margin
+  end
+  def tr
+        "|<tr>
+         |  #{yield}
+         |</tr>\n".margin
+  end
+end # module HTML
+class SpecID::Precision
+  include SpecID::Precision::PlotHelper
+  ###########################################################
+  # GLOBAL SETTINGS:
+  DEF_PREFIX = "INV_"
+  DATA_PREC = 4   # decimal places of precision for ppv data
+  STDOUT_JTPLOT_BASE = "ppv"  # if there is no outfile
+  ###########################################################
+  include SpecID::Precision::HTML
+  ## returns an html string
+  def precision(argv)
+    opt = parse_args(argv)
+    files = argv.to_a
+    out_string = prefix_as_decoy(files, opt)
+    [out_string, opt]
+  end
+  def run_cmd_line(argv)
+    output_string, opt, file_as_decoy = precision(argv)
+    if file_as_decoy
+      puts output_string
+    else
+      ## open file and write to it..
+      if opt.o == 'STDOUT'
+        print output_string
+      else
+        File.open(opt.o,'w') do |fh| fh.print output_string end
+      end
+    end
+  end
+  # returns the outfile with no extension
+  def outfile_noext(opt)
+    if opt == 'STDOUT'
+      "#{STDOUT_JTPLOT_BASE}"
+    else
+      opt.sub(/#{Regexp.escape(File.extname(opt))}$/, '')
+    end
+  end
+  def file_noext(file)
+    file.sub(/#{Regexp.escape(File.extname(file))}$/, '')
+  end
+  def parse_args(argv)
+    opt = OpenStruct.new
+    opt.o = 'STDOUT'
+    opts = OptionParser.new do |op|
+      op.banner = "Usage: #{File.basename(__FILE__)} [options] bioworks.xml|proph-prot.xml ..."
+      op.separator ""
+      op.separator "Abbreviations and Definitions:"
+      op.separator "  TP = True Positives"
+      op.separator "  FP = False Positives"
+      op.separator "  Precision = Positive Predictive Value = [TP/(TP+FP)]"
+      op.separator ""
+      op.separator "Output: "
+      op.separator "  1. Decoy as separate search: PPV to STDOUT"
+      op.separator "  2. Decoy proteins from concatenated database: '.html'"
+      op.separator ""
+      op.separator "Options:"
+      op.on("-f", "--fp_data <prefix_or_file>", "PREFIX -or- decoy FILE") {|v| opt.f = v }
+      op.separator ""
+      op.separator "        If searched with a concatenated DB, give a PREFIX to decoy proteins."
+      op.separator "        If files have different prefixes, separate with commas."
+      op.separator "        If searched with a separate decoy DB, give the FILE name of decoy data"
+      op.separator ""
+      ## NOT YET FUNCTIONAL: op.on("-e", "--peptides", "do peptides instead of proteins")
+      op.separator ""
+      op.on("-o", "--outfile <file>", "write output to file (def: #{opt.o})") {|v| opt.o = v}
+      op.on("-a", "--area", "output area under the curve instead of the plot") {|v| opt.a = v}
+      op.on("-j", "--plot_file", "output to_plot file") {|v| opt.j = v}
+      op.on_tail("
+Example:
+  For a search on a concatenated database where the decoy proteins have
+  been flagged with the prefix 'INV_' for both Bioworks and ProteinProphet
+  output:
+    #{File.basename(__FILE__)} -f INV_ bioworks.xml proph-prot.xml
+  ")
+    end
+    opts.parse!(argv)
+    if argv.size < 1
+      puts opts
+      exit
+    end
+    opt
+  end
+  # takes a comma separated list and extends the last to create an array of
+  # desired size
+  def prefixes(arg, desired_size)
+    arg_arr = arg.split(',')
+    new_arr = []
+    last_arg = arg_arr[0]
+    desired_size.times do |i|
+      if arg_arr[i]
+        new_arr[i] = arg_arr[i]
+        last_arg = new_arr[i]
+      else
+        new_arr[i] = last_arg
+      end
+    end
+    new_arr
+  end
+  ## collapses arrays to one level deep so we can sync them up
+  def arrays_to_one_level_deep(all_arrs)
+    mostly_flat = []
+    all_arrs.each do |per_file|
+      per_file.each do |per_style|
+        mostly_flat << per_style[0]
+        mostly_flat << per_style[1]
+      end
+    end
+    mostly_flat
+  end
+  # prints rows and th for the data
+  def table_cells(all_arrs, key)
+    ## columns specific headings:
+    all_string = ""
+    all_string << tr do
+      line = ""
+      key.each do |per_file|
+        per_file.each do |per_ds|
+          line << "<th class=\"small\">#{per_ds[1][0]}</th><th class=\"small\">#{per_ds[1][1]}</th>"
+        end
+      end
+      line
+    end
+    mostly_flat = arrays_to_one_level_deep(all_arrs)
+    SyncEnumerator.new(*mostly_flat).each do |row|
+      all_string << tr do
+        string = row.map {|it|
+          sty="%d"
+          if it.class == Float ; sty="%.#{DATA_PREC}f" end
+          td{ sprintf(sty,it)}
+        }.join
+      end
+    end
+    all_string
+  end
+  def html_table_output(all_arrs, key, files, filename_noext)
+    num_datasets_per_file = all_arrs.first.size
+    num_cols_per_dataset = 2
+    big_colspan = num_datasets_per_file * num_cols_per_dataset
+    output = table do
+      tr do
+        files.map do |file|
+        "<th colspan=\"#{big_colspan}\">#{file}</th>"
+        end.join
+      end +
+        tr do
+        key.map do |arr|
+          arr.map do |ds|
+        "<th colspan=\"2\">#{ds.first}</th>"
+          end
+        end
+        end +
+          table_cells(all_arrs, key)
+    end
+    "<div id=\"tp_table\">" + output + "</div>"
+  end
+  def y_axis_label(key)
+    ## We only take the keys for the first file, as it's assumed that the major
+    ## labels will be identical for all of them
+    labels = key.first.map {|tp| tp.first }
+    labels.join "  |  "
+  end
+  # escapes any ' chars
+  def escape_to_gnuplot(string)
+    # long way, but it works.
+    new_string = ""
+    string.split(//).each do |chr|
+      if chr == "'" ; new_string << "\\" end
+      new_string << chr
+    end
+    new_string
+  end
+  def prefix_as_decoy(files, opt)
+    $stderr.puts "using prefix #{opt.f} ..."
+    if opt.f
+      prefix_arr = prefixes(opt.f, files.size)
+    end
+    all_arrs = []
+    key = []
+    out_noext = outfile_noext(opt.o)
+    files.each_with_index do |file,i|
+      all_arrs[i] = []
+      key[i] = []
+      sp = SpecID.new(file)
+      #headers = ["#{file_noext(file)} Precision [TP/(TP+FP)]", "#{file_noext(file)} FPR [FP/(FP+TP)]"]
+      if opt.f
+        (num_hits, ppv) = sp.num_hits_and_ppv_for_prob(prefix_arr[i])
+        all_arrs[i] << [num_hits,ppv]
+        key[i] << ["Precision",  ["#TP", "Prec = TP/(TP+FP)"]]
+      else
+        ## These are just from protein prophet probabilities:
+        (num_hits, ppv) = sp.num_hits_and_ppv_for_protein_prophet_probabilities
+        all_arrs[i] << [num_hits,ppv]
+        key[i] << ["Precision",  ["#TP", "Prec = TP/(TP+FP)"]]
+      end
+    end
+    string = ''
+    if opt.a
+      roc = ROC.new
+      #string << "***********************************************************\n"
+      #string << "AREA UNDER CURVE:\n"
+      key.each_with_index do |file,i|
+        string << "#{files[i]} (area under curve)\n"
+        key[i].each_index do |j|
+          string << "#{key[i][j][0]} [#{ key[i][j][1]}]:\t"
+          num_hits = all_arrs[i][j][0]
+          oth = all_arrs[i][j][1]
+          string << roc.area_under_curve(num_hits, oth).to_s << "\n"
+        end
+      end
+      #string << "***********************************************************\n"
+    else
+      if opt.j
+        create_to_plot_file(all_arrs, key, files, out_noext)
+      end
+      string = html do
+        header +
+          body do
+          plot_figure(all_arrs, key, files, out_noext) +
+            html_table_output(all_arrs, key, files, out_noext)
+          end
+      end
+    end
+    string
+  end
+end # class SpecID

data/lib/spec_id/proph.rb CHANGED Viewed

@@ -14,7 +14,7 @@ class Proph
 class Parser
   def root_el(file)
-    XMLTree.parse_file(file)
+    AXML.parse_file(file)
   end
 end
@@ -275,7 +275,7 @@ class Pep::Parser < Parser
       ## file from peptideAtlas:
       search_result_regex1 = /<spectrum_query spectrum="(.*\.\d+\.\d+\.\d)".* assumed_charge="(\d)"/o
-      search_result_regex2 = /<search_result spectrum="(.*\.\d+\.\d+\.\d)".* assumed_charge="(\d)"/o
+      search_result_regex2 = /<search_result sxpectrum="(.*\.\d+\.\d+\.\d)".* assumed_charge="(\d)"/o
       search_hit_regex = /<search_hit .*peptide="(\w+)" /o
       peptide_h = {}