ulla 0.9.9.1 → 0.9.9.2

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 40c1f0642b5169fc2e4e54c12c2e7b88e06a9f7f
+  data.tar.gz: 4ac3ec4031a177629ae72cf5de1803e3749bbf66
+SHA512:
+  metadata.gz: 050a3eb74dc176396b8ef9d1f10bee85ce227e9694245776d2c0b2de92595ed919697fd79176e25c0ebeb203f1a5cc689f15f3ad6580609c8d0839b944627550
+  data.tar.gz: d05f5de53ed2ac6552af88153c932f9947d47db5dfdcbe9ef09547f54a924ff8eff1fa8cee2dbde4101f40064e5f0299208977cfff80ce239b930d9b3bcc33a4

data/History.txt

@@ -1,7 +1,7 @@
 == 0.9.9.1 26/08/2009
 
 * Removed dependency on the Ruby Facets library
-* Total -> total for compativility with melody
+* Total -> total for compatibility with melody
 
 == 0.9.9 09/08/2009
 

data/Manifest.txt

@@ -18,6 +18,10 @@ lib/ulla/environment_class_hash.rb
 lib/ulla/environment_feature.rb
 lib/ulla/environment_feature_array.rb
 lib/ulla/heatmap_array.rb
+lib/ulla/esst.rb
+lib/ulla/essts.rb
+lib/ulla/joy_tem.rb
+lib/ulla/sequence.rb
 script/console
 script/destroy
 script/generate

data/PostInstall.txt

@@ -1,5 +1,5 @@
 HELP/OPTIONS: ulla -h
 
-For more information on ulla, see http://ulla.rubyforge.org
+For more information on ulla, see http://github.com/semin/ulla.
 
 

data/README.rdoc

@@ -25,7 +25,6 @@ Following RubyGems will be automatically installed if you have rubygems installe
 
 * narray (http://narray.rubyforge.org)
 * bio (http://bioruby.open-bio.org)
-* Active Support (http://as.rubyonrails.org)
 * RMagick (http://rmagick.rubyforge.org)
 
 
@@ -36,7 +35,7 @@ Following RubyGems will be automatically installed if you have rubygems installe
 
 == Basic Usage
 
-It's pretty much the same as Kenji's subst (http://www-cryst.bioc.cam.ac.uk/~kenji/subst/), so in most cases, you can swap 'subst' with 'ulla'.
+It's pretty much the same as Kenji's subst (http://mordred.bioc.cam.ac.uk/~kenji/subst/), so in most cases, you can swap 'subst' with 'ulla'.
 
     ~user $ ulla -l TEMLIST-file -c classdef.dat
                 or
@@ -98,7 +97,7 @@ It's pretty much the same as Kenji's subst (http://www-cryst.bioc.cam.ac.uk/~ken
 
 == Usage
 
-1. Prepare an environmental class definition file. For more details, please check this notes (http://www-cryst.bioc.cam.ac.uk/~kenji/subst/NOTES). You can download a sample environmental class definition file from http://www-cryst.bioc.cam.ac.uk/~kenji/subst/classdef.dat 
+1. Prepare an environmental class definition file. For more details, please check this notes (http://mordred.bioc.cam.ac.uk/~kenji/subst/NOTES). You can download a sample environmental class definition file from http://mordred.bioc.cam.ac.uk/~kenji/subst/classdef.dat 
 
         ~user $ cat classdef.dat
         #
@@ -108,7 +107,7 @@ It's pretty much the same as Kenji's subst (http://www-cryst.bioc.cam.ac.uk/~ken
         secondary structure and phi angle;HEPC;HEPC;T;F
         solvent accessibility;TF;Aa;F;F
 
-2. Prepare structural alignments and their annotations of above environmental classes in PIR format. You can download sample alignments from http://www-cryst.bioc.cam.ac.uk/~kenji/subst/alltem-allmask.tar.gz
+2. Prepare structural alignments and their annotations of above environmental classes in PIR format. You can download sample alignments from http://mordred.bioc.cam.ac.uk/~kenji/subst/alltem-allmask.tar.gz or from http://www-cryst.bioc.cam.ac.uk/ESST/
 
         ~user $ cat sample1.tem
         >P1;1mnma
@@ -167,7 +166,7 @@ It's pretty much the same as Kenji's subst (http://www-cryst.bioc.cam.ac.uk/~ken
 
 9. In case positions are masked with the character 'X' in any environmental features, all mutations from/to the position will be excluded from substitution counts.
 
-10. Then, it will produce a file containing all the matrices, which will look like the one below. For more details, please check this notes (http://www-cryst.bioc.cam.ac.uk/~kenji/subst/NOTES).
+10. Then, it will produce a file containing all the matrices, which will look like the one below. For more details, please check this notes (http://mordred.bioc.cam.ac.uk/~kenji/subst/NOTES).
 
         # Environment-specific amino acid substitution matrices
         # Creator: ulla version 0.0.5
@@ -226,7 +225,7 @@ It's pretty much the same as Kenji's subst (http://www-cryst.bioc.cam.ac.uk/~ken
 
     which will look like this,
 
-    http://www-cryst.bioc.cam.ac.uk/~semin/images/0.HA.png
+    http://mordred.bioc.cam.ac.uk/~semin/images/0.HA.png
 
 12. To generate one big figure, 'myheatmaps.gif' containing all the heat maps (4 maps in a row),
 
@@ -234,7 +233,7 @@ It's pretty much the same as Kenji's subst (http://www-cryst.bioc.cam.ac.uk/~ken
 
     which will look like this,
 
-    http://www-cryst.bioc.cam.ac.uk/~semin/images/myheatmaps.gif
+    http://mordred.bioc.cam.ac.uk/~semin/images/myheatmaps.gif
 
 == Repository
 

data/Rakefile

@@ -2,6 +2,7 @@ require 'rubygems'
 gem 'hoe', '>= 2.1.0'
 require 'hoe'
 require 'fileutils'
+require './lib/ulla.rb'
 
 Hoe.plugin :newgem
 # Hoe.plugin :website
@@ -12,7 +13,6 @@ Hoe.plugin :newgem
 $hoe = Hoe.spec 'ulla' do
   self.developer 'Semin Lee', 'seminlee@gmail.com'
   self.post_install_message = 'PostInstall.txt' # TODO remove if post-install message not required
-  self.rubyforge_name       = self.name # TODO this is default value
   self.extra_deps           = [
     ['narray',        '>= 0.5.9.5'],
     ['bio',           '>= 1.2.1'],
@@ -21,4 +21,4 @@ $hoe = Hoe.spec 'ulla' do
 end
 
 require 'newgem/tasks'
-Dir['tasks/**/*.rake'].each { |t| load t }
+Dir['tasks/*.rake'].each { |t| load t }

data/lib/narray_extensions.rb

@@ -1,6 +1,3 @@
-require 'rubygems'
-require 'narray'
-
 module NArrayExtensions
 
   def pretty_string(options={})

data/lib/nmatrix_extensions.rb

@@ -1,14 +1,3 @@
-require 'rubygems'
-require 'narray'
-
-begin
-  require 'rvg/rvg'
-  include Magick
-rescue Exception => e
-  $logger.warn "#{e.to_s.chomp} For this reason, heat maps cannot be generated."
-  $no_rmagick = true
-end
-
 module NMatrixExtensions
 
   def pretty_string(options={})

data/lib/ulla.rb

@@ -1,6 +1,50 @@
 $:.unshift(File.dirname(__FILE__)) unless
   $:.include?(File.dirname(__FILE__)) || $:.include?(File.expand_path(File.dirname(__FILE__)))
 
+require 'bio'
+require 'set'
+require 'logger'
+require 'narray'
+require 'rubygems'
+require 'bio'
+require 'set'
+require 'inline'
+require 'narray'
+require 'logger'
+require 'narray'
+require 'stringio'
+require 'pathname'
+require 'getoptlong'
+require 'fork_manager'
+require 'facets/enumerable'
+
+begin
+  require 'rvg/rvg'
+  include Magick
+rescue Exception => e
+  $logger.warn "#{e.to_s.chomp} For this reason, heat maps cannot be generated."
+  $no_rmagick = true
+end
+
+require_relative 'math_extensions'
+require_relative 'array_extensions'
+require_relative 'string_extensions'
+require_relative 'narray_extensions'
+require_relative 'nmatrix_extensions'
+
+require_relative 'ulla/esst'
+require_relative 'ulla/essts'
+require_relative 'ulla/joy_tem'
+require_relative 'ulla/sequence'
+require_relative 'ulla/heatmap_array'
+require_relative 'ulla/environment'
+require_relative 'ulla/environment_class_hash'
+require_relative 'ulla/environment_feature'
+require_relative 'ulla/environment_feature_array'
+
 module Ulla
-  VERSION = '0.9.9.1'
+  VERSION = '0.9.9.2'
+
+  $logger       = Logger.new(STDOUT)
+  $logger.level = Logger::WARN
 end

data/lib/ulla/cli.rb

@@ -1,15 +1,86 @@
-require 'rubygems'
-require 'getoptlong'
-require 'logger'
-require 'narray'
-require 'bio'
-require 'set'
-
 # This is a module for an actual command line interpreter for Ulla
 # ---
 # Copyright (C) 2008-9 Semin Lee
 module Ulla
   class CLI
+
+    # Calculate PID between two sequences
+    #
+    # :call-seq:
+    #   Ulla::CLI::calculate_pid(seq1, seq2, unit) -> Float
+    #
+    def self.calculate_pid_rb(seq1, seq2, unit)
+      aas1  = seq1.scan(/\S{#{unit}}/)
+      aas2  = seq2.scan(/\S{#{unit}}/)
+      gap   = ($gap || '-') * unit
+      align = 0 # no. of aligned columns
+      ident = 0 # no. of identical columns
+      intgp = 0 # no. of internal gaps
+
+      if (aas1.size != aas2.size)
+        $logger.error "Cannot calculate PID between unaligned sequences"
+        $logger.error seq1, seq2
+        exit 1
+      end
+
+      (0...aas1.size).each do |i|
+        if (aas1[i] != gap) && (aas2[i] != gap)
+          align += 1
+          if aas1[i] == aas2[i]
+            ident += 1
+          end
+        elsif (((aas1[i] == gap) && (aas2[i] != gap)) ||
+                ((aas1[i] != gap) && (aas2[i] == gap)))
+          intgp += 1
+        end
+      end
+
+      100.0 * ident / (align + intgp)
+    end
+
+
+    inline(:C) do |builder|
+      builder.add_compile_flags '-x c++', '-lstdc++'
+      builder.c_singleton %q{
+        static VALUE calculate_pid_cpp(VALUE seq1, VALUE seq2, VALUE unit) {
+          VALUE re      = rb_str_plus(rb_str_plus(rb_str_new2("\\\\S{"), rb_funcall(unit, rb_intern("to_s"), 0)), rb_str_new2("}"));
+          VALUE aas1    = rb_funcall(seq1, rb_intern("scan"), 1, rb_reg_new_str(re, 0));
+          VALUE aas2    = rb_funcall(seq2, rb_intern("scan"), 1, rb_reg_new_str(re, 0));
+          //VALUE aas1    = rb_funcall(seq1, rb_intern("split"), 1, rb_str_new2(""));
+          //VALUE aas2    = rb_funcall(seq2, rb_intern("split"), 1, rb_str_new2(""));
+          VALUE *aas1_p = RARRAY_PTR(aas1);
+          VALUE *aas2_p = RARRAY_PTR(aas2);
+          VALUE gap     = rb_str_new2("-");
+          long  len1    = RARRAY_LEN(aas1);
+          //long  len2     = RARRAY_LEN(aas2);
+          double align  = 0.0;
+          double ident  = 0.0;
+          double intgp  = 0.0;
+
+          for (long i = 0; i < len1; i++) {
+            if ((rb_str_equal(aas1_p[i], gap) == Qfalse) && (rb_str_equal(aas2_p[i], gap) == Qfalse)) {
+              align += 1.0;
+              if (rb_str_equal(aas1_p[i], aas2_p[i]) == Qtrue) {
+                ident += 1.0;
+              }
+            } else if (((rb_str_equal(aas1_p[i], gap) == Qtrue) && (rb_str_equal(aas2_p[i], gap) == Qfalse)) ||
+                        ((rb_str_equal(aas1_p[i], gap) == Qfalse) && (rb_str_equal(aas2_p[i], gap) == Qtrue))) {
+              intgp += 1.0;
+            }
+          }
+          return DBL2NUM(100.0 * ident / (align + intgp));
+        }
+      }
+    end
+
+    def self.calculate_pid(seq1, seq2, unit)
+      begin
+        self.calculate_pid_cpp(seq1, seq2, unit)
+      rescue
+        self.calculate_pid_rb(seq1, seq2, unit)
+      end
+    end
+
     class << self
 
       # :nodoc:
@@ -38,7 +109,7 @@ Options:
     --tem-file (-f) FILE: a tem file
     --tem-list (-l) FILE: a list for tem files
     --classdef (-c) FILE: a file for the defintion of environmental class
-                          if no definition file provided, --cys (-y) 2 and --nosmooth options automatcially applied
+                          if no definition file provided, --cys (-y) 2 and --nosmooth options applied
     --outfile (-o) FILE: output filename (default 'allmat.dat')
     --weight (-w) INTEGER: clustering level (PID) for the BLOSUM-like weighting (default: 60)
     --noweight: calculate substitution counts with no weights
@@ -58,13 +129,13 @@ Options:
         0 for raw counts (no smoothing performed)
         1 for probabilities
         2 for log-odds (default)
-    --noroundoff: do not round off log odds ratio
-    --scale INTEGER: log-odds matrices in 1/n bit units (default 3)
-    --sigma DOUBLE: change the sigma value for smoothing (default 5.0)
+    --noroundoff: do not round off log-odds ratio
+    --scale INTEGER: log-odds matrices in 1/n bit units (default: 3)
+    --sigma DOUBLE: change the sigma value for smoothing (default: 5.0)
     --autosigma: automatically adjust the sigma value for smoothing
-    --add DOUBLE: add this value to raw counts when deriving log-odds without smoothing (default 0)
-    --pidmin DOUBLE: count substitutions only for pairs with PID equal to or greater than this value (default none)
-    --pidmax DOUBLE: count substitutions only for pairs with PID smaller than this value (default none)
+    --add DOUBLE: add this value to raw counts when deriving log-odds without smoothing (default: 0)
+    --pidmin DOUBLE: count substitutions only for pairs with PID equal to or greater than this value
+    --pidmax DOUBLE: count substitutions only for pairs with PID smaller than this value
     --heatmap INTEGER:
         0 create a heat map file for each substitution table
         1 create one big file containing all heat maps from substitution tables
@@ -91,35 +162,6 @@ Options:
         puts (verbose ? usage + options : usage)
       end
 
-      # Calculate PID between two sequences
-      #
-      # :call-seq:
-      #   Ulla::CLI::calculate_pid(seq1, seq2) -> Float
-      #
-      def calculate_pid(seq1, seq2, unit)
-        aas1  = seq1.scan(/\w{#{unit}}/)
-        aas2  = seq2.scan(/\w{#{unit}}/)
-        cols  = aas1.zip(aas2)
-        gap   = ($gap || '-') * unit
-        align = 0 # no. of aligned columns
-        ident = 0 # no. of identical columns
-        intgp = 0 # no. of internal gaps
-
-        cols.each do |col|
-          if (col[0] != gap) && (col[1] != gap)
-            align += 1
-            if col[0] == col[1]
-              ident += 1
-            end
-          elsif (((col[0] == gap) && (col[1] != gap)) ||
-                 ((col[0] != gap) && (col[1] == gap)))
-            intgp += 1
-          end
-        end
-
-        pid = 100.0 * ident.to_f / (align + intgp)
-      end
-
       # :nodoc:
       def execute(arguments=[])
         #
@@ -152,9 +194,6 @@ Options:
         # Global variables and their default values
         #
 
-        $logger       = Logger.new(STDOUT)
-        $logger.level = Logger::WARN
-
         # default set of 21 amino acids including J (Cysteine, the free thiol form)
         $amino_acids    = 'ACDEFGHIKLMNPQRSTVWYJ'.split('')
         $gap            = '-'
@@ -179,7 +218,6 @@ Options:
         $scale          = 3
         $pidmin         = nil
         $pidmax         = nil
-        $scale          = 3
         $add            = nil
         $cys            = 0
         $targetenv      = false
@@ -233,6 +271,9 @@ Options:
           [ '--noroundoff',         GetoptLong::NO_ARGUMENT ],
           [ '--sigma',              GetoptLong::REQUIRED_ARGUMENT ],
           [ '--autosigma',          GetoptLong::NO_ARGUMENT ],
+          [ '--scale',              GetoptLong::REQUIRED_ARGUMENT ],
+          [ '--pidmax',             GetoptLong::REQUIRED_ARGUMENT ],
+          [ '--pidmin',             GetoptLong::REQUIRED_ARGUMENT ],
           [ '--add',                GetoptLong::REQUIRED_ARGUMENT ],
           [ '--heatmap',            GetoptLong::REQUIRED_ARGUMENT ],
           [ '--heatmap-stem',       GetoptLong::REQUIRED_ARGUMENT ],
@@ -297,7 +338,7 @@ Options:
             when '--penv'
               warn "--penv option is not supported."
               exit 1
-              $penv         = true
+              #$penv         = true
             when '--heatmap'
               $heatmap      = case arg.to_i
                               when (0..2) then arg.to_i
@@ -365,19 +406,6 @@ Options:
           warn "Cannot find environment class definition file, #{$classdef}"
           exit 1
         end
-
-	require 'math_extensions'
-	require 'array_extensions'
-	require 'string_extensions'
-	require 'narray_extensions'
-	require 'nmatrix_extensions'
-
-	require 'ulla/environment'
-	require 'ulla/environment_class_hash'
-	require 'ulla/environment_feature'
-	require 'ulla/environment_feature_array'
-	require 'ulla/heatmap_array'
-
         #
         # Part 2 END
         #
@@ -425,15 +453,18 @@ Options:
               next
             elsif (env_ftr = line.split(/;/)).length == 5
               $logger.info "An environment feature, #{line} detected."
+
               if env_ftr[-1] == 'T'
                 # skip silenced environment feature
                 $logger.warn "The environment feature, #{line} silent."
                 next
               end
+
               if env_ftr[-2] == 'T'
                 $cst_features << env_index
                 $logger.warn "The environment feature, #{line} constrained."
               end
+
               $env_features << EnvironmentFeature.new(env_ftr[0],
                                                       env_ftr[1].split(''),
                                                       env_ftr[2].split(''),
@@ -571,7 +602,7 @@ Options:
                     seq2 = seq2.split('').each_with_index.map { |aa, pos| aa == $gap ? $ext_gap : env_labels[id2][pos] }.join
                   end
 
-                  pid = calculate_pid(seq1, seq2, $col_size)
+                  pid = calculate_pid_cpp(seq1, seq2, $col_size)
                   s1  = seq1.scan(/\S{#{$col_size}}/)
                   s2  = seq2.scan(/\S{#{$col_size}}/)
 
@@ -610,8 +641,10 @@ Options:
                       next
                     end
 
-                    aa1       = (disulphide.has_key?(id1) && (disulphide[id1][pos] == 'F') && (aa1[0].chr == 'C') && ($cys != 2)) ? 'J' + aa1[1..-1] : aa1
-                    aa2       = (disulphide.has_key?(id2) && (disulphide[id2][pos] == 'F') && (aa2[0].chr == 'C') && ($cys != 2)) ? 'J' + aa2[1..-1] : aa2
+                    #aa1       = (disulphide.has_key?(id1) && (disulphide[id1][pos] == 'F') && (aa1[0].chr == 'C') && ($cys != 2)) ? 'J' + aa1[1..-1] : aa1
+                    #aa2       = (disulphide.has_key?(id2) && (disulphide[id2][pos] == 'F') && (aa2[0].chr == 'C') && ($cys != 2)) ? 'J' + aa2[1..-1] : aa2
+                    aa1       = (aa1[0].chr == 'C' && (!disulphide.has_key?(id1) || disulphide[id1][pos] == 'F') && $cys != 2) ? 'J' + aa1[1..-1] : aa1
+                    aa2       = (aa2[0].chr == 'C' && (!disulphide.has_key?(id2) || disulphide[id2][pos] == 'F') && $cys != 2) ? 'J' + aa2[1..-1] : aa2
                     env_label = $environment == 1 ? aa1 + '-' + aa2[1..-1] : env_labels[id1][pos]
 
                     if $cst_features.empty?
@@ -648,7 +681,7 @@ Options:
               ali = ext_ali
             end
 
-            # a loop for single linkage clustering
+            # loop for single linkage clustering
             begin
               continue = false
               0.upto(clusters.size - 2) do |i|
@@ -657,7 +690,7 @@ Options:
                   found = false
                   clusters[i].each do |c1|
                     clusters[j].each do |c2|
-                      if calculate_pid(ali[c1], ali[c2], $col_size) >= $weight
+                      if calculate_pid_cpp(ali[c1], ali[c2], $col_size) >= $weight
                         indexes << j
                         found = true
                         break
@@ -694,12 +727,12 @@ Options:
                   seq1.each_with_index do |aa1, pos|
                     aa2 = seq2[pos]
 
-                    if env_labels[id1][pos].include?('X')
+                    if env_labels.has_key?(id1) && env_labels[id1][pos].include?('X')
                       $logger.debug "All substitutions from #{id1}-#{pos}-#{aa1[0].chr} are masked."
                       next
                     end
 
-                    if env_labels[id2][pos].include?('X')
+                    if env_labels.has_key?(id2) && env_labels[id2][pos].include?('X')
                       $logger.debug "All substitutions to #{id2}-#{pos}-#{aa2[0].chr} are masked."
                       next
                     end
@@ -714,18 +747,21 @@ Options:
                       next
                     end
 
-                    aa1         = (disulphide.has_key?(id1) && (disulphide[id1][pos] == 'F') && (aa1[0].chr == 'C') && ($cys != 2)) ? 'J' + aa1[1..-1] : aa1
-                    aa2         = (disulphide.has_key?(id2) && (disulphide[id2][pos] == 'F') && (aa2[0].chr == 'C') && ($cys != 2)) ? 'J' + aa2[1..-1] : aa2
-                    cnt1        = 1.0 / cluster1.size.to_f
-                    cnt2        = 1.0 / cluster2.size.to_f
+                    #aa1         = (disulphide.has_key?(id1) && (disulphide[id1][pos] == 'F') && (aa1[0].chr == 'C') && ($cys != 2)) ? 'J' + aa1[1..-1] : aa1
+                    #aa2         = (disulphide.has_key?(id2) && (disulphide[id2][pos] == 'F') && (aa2[0].chr == 'C') && ($cys != 2)) ? 'J' + aa2[1..-1] : aa2
+                    #aa1         = (aa1[0].chr == 'C' && (!disulphide.has_key?(id1) || disulphide[id1][pos] == 'F') && $cys != 2) ? 'J' + aa1[1..-1] : aa1
+                    #aa2         = (aa2[0].chr == 'C' && (!disulphide.has_key?(id2) || disulphide[id2][pos] == 'F') && $cys != 2) ? 'J' + aa2[1..-1] : aa2
+                    cnt1        = 1.0 / cluster1.size
+                    cnt2        = 1.0 / cluster2.size
                     jnt_cnt     = cnt1 * cnt2
                     env_label1  = $environment == 1 ? aa1 + '-' + aa2[1..-1] : env_labels[id1][pos]
                     env_label2  = $environment == 1 ? aa2 + '-' + aa1[1..-1] : env_labels[id2][pos]
 
                     if $cst_features.empty?
-                      $env_classes[env_label1].increase_residue_count(aa2[0].chr, jnt_cnt) #rescue $logger.error "Something wrong with #{tem_file}-#{id2}-#{pos}-#{aa2}-#{env_label2}"
-                      $env_classes[env_label2].increase_residue_count(aa1[0].chr, jnt_cnt) #rescue $logger.error "Something wrong with #{tem_file}-#{id2}-#{pos}-#{aa2}-#{env_label2}"
-                    elsif (env_labels[id1][pos].split('').values_at(*$cst_features) == env_labels[id2][pos].split('').values_at(*$cst_features))
+                      $env_classes[env_label1].increase_residue_count(aa2[0].chr, jnt_cnt)
+                      $env_classes[env_label2].increase_residue_count(aa1[0].chr, jnt_cnt)
+                    elsif (env_labels[id1][pos].split('').values_at(*$cst_features) ==
+                           env_labels[id2][pos].split('').values_at(*$cst_features))
                       $env_classes[env_label1].increase_residue_count(aa2[0].chr, jnt_cnt)
                       $env_classes[env_label2].increase_residue_count(aa1[0].chr, jnt_cnt)
                     else
@@ -735,8 +771,71 @@ Options:
 
                     $aa_tot_cnt.has_key?(aa1) ? $aa_tot_cnt[aa1] += cnt1 : $aa_tot_cnt[aa1] = cnt1
                     $aa_tot_cnt.has_key?(aa2) ? $aa_tot_cnt[aa2] += cnt2 : $aa_tot_cnt[aa2] = cnt2
-                    $aa_mut_cnt.has_key?(aa1) ? $aa_mut_cnt[aa1] += cnt1 : $aa_mut_cnt[aa1] = cnt1 if aa1 == aa2
-                    $aa_mut_cnt.has_key?(aa2) ? $aa_mut_cnt[aa2] += cnt2 : $aa_mut_cnt[aa2] = cnt2 if aa1 == aa2
+                    ($aa_mut_cnt.has_key?(aa1) ? $aa_mut_cnt[aa1] += cnt1 : $aa_mut_cnt[aa1] = cnt1) if aa1 != aa2
+                    ($aa_mut_cnt.has_key?(aa2) ? $aa_mut_cnt[aa2] += cnt2 : $aa_mut_cnt[aa2] = cnt2) if aa1 != aa2
+
+                    #if $cst_features.empty?
+                      #if $env_classes.has_key?(env_label1)
+                        #if $env_classes.has_key?(env_label2)
+                          #$env_classes[env_label1].increase_residue_count(aa2[0].chr, jnt_cnt)
+                        #else
+                          #if (aa1 == 'C' && aa2 == 'J')
+                            #$env_classes[env_label1].increase_residue_count('C', jnt_cnt)
+                          #else
+                            #$env_classes[env_label1].increase_residue_count(aa2, jnt_cnt)
+                          #end
+                        #end
+                      #end
+                      #if $env_classes.has_key?(env_label2)
+                        #if $env_classes.has_key?(env_label1)
+                          #$env_classes[env_label2].increase_residue_count(aa1[0].chr, jnt_cnt)
+                        #else
+                          #if (aa2 == 'C' && aa1 == 'J')
+                            #$env_classes[env_label2].increase_residue_count('C', jnt_cnt)
+                          #else
+                            #$env_classes[env_label2].increase_residue_count(aa1, jnt_cnt)
+                          #end
+                        #end
+                      #end
+                    #elsif (env_labels[id1][pos].split('').values_at(*$cst_features) == env_labels[id2][pos].split('').values_at(*$cst_features))
+                      #$env_classes[env_label1].increase_residue_count(aa2[0].chr, jnt_cnt)
+                      #$env_classes[env_label2].increase_residue_count(aa1[0].chr, jnt_cnt)
+                    #else
+                      #$logger.debug "Skipped #{id1}-#{pos}-#{aa1[0].chr} and #{id2}-#{pos}-#{aa2[0].chr} having different symbols for constrained environment features each other."
+                      #next
+                    #end
+
+                    #if $env_classes.has_key?(env_label1)
+                      #$aa_tot_cnt.has_key?(aa1) ? $aa_tot_cnt[aa1] += cnt1 : $aa_tot_cnt[aa1] = cnt1
+
+                      #if $env_classes.has_key?(env_label2)
+                        #if aa1[0].chr != aa2[0].chr
+                          #$aa_mut_cnt.has_key?(aa1) ? $aa_mut_cnt[aa1] += cnt1 : $aa_mut_cnt[aa1] = cnt1
+                        #end
+                      #else
+                        #if (aa1[0].chr != aa2)
+                          #unless (aa1[0].chr == 'C' && aa2 == 'J') || (aa1[0].chr == 'J' && aa2 == 'C')
+                            #$aa_mut_cnt.has_key?(aa1) ? $aa_mut_cnt[aa1] += cnt1 : $aa_mut_cnt[aa1] = cnt1
+                          #end
+                        #end
+                      #end
+                    #end
+
+                    #if $env_classes.has_key?(env_label2)
+                      #$aa_tot_cnt.has_key?(aa2) ? $aa_tot_cnt[aa2] += cnt2 : $aa_tot_cnt[aa2] = cnt2
+
+                      #if $env_classes.has_key?(env_label1)
+                        #if aa1[0].chr != aa2[0].chr
+                          #$aa_mut_cnt.has_key?(aa2) ? $aa_mut_cnt[aa2] += cnt2 : $aa_mut_cnt[aa2] = cnt2
+                        #end
+                      #else
+                        #if (aa1 != aa2[0].chr)
+                          #unless (aa1 == 'J' && aa2[0].chr == 'C') || (aa1 == 'C' && aa2[0].chr == 'J')
+                            #$aa_mut_cnt.has_key?(aa2) ? $aa_mut_cnt[aa2] += cnt2 : $aa_mut_cnt[aa2] = cnt2
+                          #end
+                        #end
+                      #end
+                    #end
 
                     $logger.debug "#{id1}-#{pos}-#{aa1[0].chr} -> #{id2}-#{pos}-#{aa2[0].chr} substitution count (#{"%.2f" % jnt_cnt}) was added to the environments class, #{env_label1}."
                     $logger.debug "#{id2}-#{pos}-#{aa2[0].chr} -> #{id1}-#{pos}-#{aa1[0].chr} substitution count (#{"%.2f" % jnt_cnt}) was added to the environments class, #{env_label2}."
@@ -748,6 +847,13 @@ Options:
           $logger.info "Analysing #{tem_file} done."
         end
 
+        $tot_aa = $aa_tot_cnt.values.sum
+
+        if $tot_aa < 1
+          $logger.warn "No amino acid substitution counted!"
+          exit 1
+        end
+
         # print out default header
         $outfh.puts <<HEADER
 # Environment-specific amino acid substitution matrices
@@ -798,18 +904,15 @@ HEADER
 
         # calculate amino acid frequencies and mutabilities, and
         # print them as default statistics in the header part
+
+        # pre-calculate ALA's mutability
         if $environment == 0
-          ala_factor  = if $aa_tot_cnt['A'] == 0
-                          0.0
-                        elsif $aa_mut_cnt['A'] == 0
-                          0.0
-                        else
-                          100.0 * $aa_tot_cnt['A'] / $aa_mut_cnt['A'].to_f
-                        end
+          ala_mutb = if $aa_tot_cnt['A'] == 0 then 0.0
+                     elsif $aa_mut_cnt['A'] == 0 then 0.0
+                     else $aa_mut_cnt['A'].to_f / $aa_tot_cnt['A']
+                     end
         end
 
-        $tot_aa = $aa_tot_cnt.values.sum
-
         $outfh.puts '#'
         $outfh.puts "# Total amino acid frequencies:\n"
 
@@ -843,8 +946,8 @@ HEADER
           end
 
           if $environment == 0
-            $aa_mutb[aa]     = ($aa_tot_cnt[aa] == 0) ? 1.0 : ($aa_mut_cnt[aa] / $aa_tot_cnt[aa].to_f)
-            $aa_rel_mutb[aa] = $aa_mutb[aa] * ala_factor
+            $aa_mutb[aa]     = ($aa_tot_cnt[aa] == 0) ? 0.0 : ($aa_mut_cnt[aa] / $aa_tot_cnt[aa].to_f)
+            $aa_rel_mutb[aa] = 100 * $aa_mutb[aa] / ala_mutb
           end
 
           $aa_tot_freq[aa] = ($aa_tot_cnt[aa] == 0) ? 0.0 : ($aa_tot_cnt[aa] / $tot_aa.to_f)
@@ -866,7 +969,7 @@ HEADER
 
           if $noweight
             if $environment == 0
-              $outfh.puts '# %-3s %9d %9d %5.2f %8d %8.4f' % columns
+              $outfh.puts '# %-3s %9d %9d %5.2f %8d %8.6f' % columns
             else
               $outfh.puts "# %-3s %-#{$env_features.size}s %9d %9d %8.4f" % columns
             end