pasv_lib 0.4.0 → 0.4.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 752ca72e197ff3f4249ba50b67cbe8ecc09a00aa8e4f4b050943437c303bbc42
-  data.tar.gz: 35c1ed983ba5643ea84e31fa3c7ca6c0ecb031597ea2d8058664725f443e336a
+  metadata.gz: 9c784b906c7e31bdcd730c366169b973d47a2d42684515bcb80bdc07a1885deb
+  data.tar.gz: 6e8d255293a94a02fcaf4cee373632c387c753db7f71b8be6515c75ea84e22f6
 SHA512:
-  metadata.gz: 5436d6c890346eb6560d6f00d4801ca5f59f61e0d517404429553b23750f4516f117a86fb8236cc1cabd7f1a07251a2ce6f41cc2df8c3210620ac0204e686e5b
-  data.tar.gz: e92b04401c6b118b7906eca7e750f42dcfe720d4d42c3d533de33d6a8f46babb6601bdeeb2ed8e80bd4bce9407e084f8a8d60c0e6215a8e254aa96ab79442511
+  metadata.gz: ed50ff3cf8856cf568bb20f38f0adcd045699244b468909e88452a5fbc930477e26618dc0e18b0a7c6e332a6f8de1a48803b44ba228812eac53b1b35e92a38a4
+  data.tar.gz: d73b1d2bcbc5b28a951017d6074b7ce5cdc7435684c86be4df4911c5d01924164606cce8c679b638206b2c951ac7506567ef97a29dee6e848d3f8f051b3f7c73

data/lib/pasv_lib/alignment.rb

@@ -3,6 +3,7 @@ require "set"
 
 module PasvLib
   module Alignment
+    # If you need to check if a residue is a gap, use this Set.
     GAP_CHARS = Set.new %w[- .]
 
     # If the overall min of the scoring matrix is < 0, then this scales it so that the overall min becomes zero.
@@ -44,7 +45,7 @@ module PasvLib
     #   klass.alignment_columns ["AA-", "A-A"] #=> [["A", "A"], ["A", "-"], ["-", "A"]]
     def alignment_columns seqs
       seqs_no_spaces = seqs.map { |seq| seq.tr " ", "" }
-      len = seqs_no_spaces.first.length
+      len            = seqs_no_spaces.first.length
 
       seqs_no_spaces.map do |seq|
         unless seq.length == len
@@ -55,6 +56,10 @@ module PasvLib
       end.transpose
     end
 
+    def gap? residue
+      GAP_CHARS.include? residue
+    end
+
     # Calculate the geometric index for an alignment.
     #
     # Basically, you change all residues to 0 and all gaps to 1.  Then you take the permutations of the sequences and then the residues and XOR each of the permutations.  Then you add it up and take averages and you'll get the score.  That is a pretty bad explanation, so see http://merenlab.org/2016/11/08/pangenomics-v2/#geometric-homogeneity-index for more information.
@@ -72,8 +77,8 @@ module PasvLib
         binary_aln = binary_aln.transpose
       end
 
-      num_rows = binary_aln.length
-      max_differences_per_row = binary_aln.first.length
+      num_rows                 = binary_aln.length
+      max_differences_per_row  = binary_aln.first.length
       num_comparisions_per_row = num_rows - 1
 
       diff_score = binary_aln.permutation(2).map do |(row1, row2)|
@@ -91,13 +96,21 @@ module PasvLib
     #
     # @return [Float] a score between 0 and 1, with 1 being very homogeneous and 1 being very heterogeneous.
     def geometric_index aln
-      by_seq_score = geometric_score aln, "sequence"
+      by_seq_score     = geometric_score aln, "sequence"
       by_residue_score = geometric_score aln, "residue"
 
       (by_seq_score + by_residue_score) / 2.0
     end
 
-    # @note Technically you could get a negative score if you don't have enough high scoring residue pairs to keep the total score above zero.  If this is the case, you're alignment probably isn't very good.
+    # Returns the similarity score for the alignment.
+    #
+    # For each colunm, each pair of residues are scored using the similarity matrix to get a points accrued / max_points.  Then all column scores are averaged.
+    #
+    # @raise PasvLib::Error if any residue is not present in the scoring matrix.
+    # @raise PasvLib::Error if seqs is empty
+    # @raise PasvLib::Error if max_points is zero.  Could happen if one of the seqs has all gaps, or no omparisons could be made.
+    #
+    # @note Technically you could get a negative score if you don't have enough high scoring residue pairs to keep the total score above zero.  If this is the case, you're alignment probably isn't very good.  Alternatively, you could use #adjust_scoring_matrix to avoid this issue....
     def similarity_score seqs, scoring_matrix = Blosum::BLOSUM62
       raise PasvLib::Error if seqs.empty?
       return 1.0 if seqs.count == 1
@@ -109,7 +122,15 @@ module PasvLib
 
       aln_cols.each do |residues|
         residues.map(&:upcase).combination(2).each do |r1, r2|
-          unless GAP_CHARS.include?(r1) || GAP_CHARS.include?(r2)
+          unless gap?(r1) || gap?(r2)
+            # Check that scoring matrix has the residues.
+            [r1, r2].each do |res|
+              unless scoring_matrix.has_key? res
+                raise PasvLib::Error, "Residue '#{res}' is missing from the " \
+                                    "scoring matrix."
+              end
+            end
+
             r1_max_score = scoring_matrix[r1].values.max
             r2_max_score = scoring_matrix[r2].values.max
             pair_max     = [r1_max_score, r2_max_score].max
@@ -139,7 +160,7 @@ module PasvLib
       aln.map do |seq|
         seq.chars.map do |char|
           # Gaps turn to 1, residues turn to 0.
-          GAP_CHARS.include?(char) ? 1 : 0
+          gap?(char) ? 1 : 0
         end
       end
     end

data/lib/pasv_lib/version.rb

@@ -1,3 +1,3 @@
 module PasvLib
-  VERSION = "0.4.0"
+  VERSION = "0.4.1"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pasv_lib
 version: !ruby/object:Gem::Version
-  version: 0.4.0
+  version: 0.4.1
 platform: ruby
 authors:
 - Ryan Moore