libssw 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: c888d3c126247f4f397fc151402fd482cdc4c105bd00163d07cc056b0e202607
-  data.tar.gz: 8717cde76e5ba26034b05c81871b43730580b09dfb610c9507b0c651e6b816a1
+  metadata.gz: 182b0d30cdf3d9a93b100b05f96d469d612b09bcfcfd2afc3cb63a4c93501d17
+  data.tar.gz: 99ec370125c707acff3a99664706ce8691f2e66e48d22a49af0362894b2c6b72
 SHA512:
-  metadata.gz: d10a0cc734b8c53bf506c97ade223269ad6a64ed356f81d08eb9368e031b068ed85c217d3e47a737a303604b57009a00c0aab94a6eea89a62382cf79dbbee326
-  data.tar.gz: f18a9eddc8455e13a58058294d80f52ed96b31949e90c25494fd48ac7d3f22df4588e86df389160fca973489db36381bdabb67a2858ceeb811e9a8ddba3249cb
+  metadata.gz: 8a49bf4924da5d12310b691f9f9335d96cdce0ef4b82ac1a38aeb6293c92d939cd17e4b0e6e99a4c427a0f4074f94db1cca91ca6c26b76a0551e558371b4aee7
+  data.tar.gz: 5af24c2124cd53f8aa54dc905172eb6f9b3d7c57b33fb87d1884ca2ab1a90acf0cb1f2ca73cb88eecda223e3f723836cca2dbae9cb0c1ac665d097993645ee19

data/README.md

@@ -1,6 +1,8 @@
 # ruby-libssw
 
 ![test](https://github.com/kojix2/ruby-libssw/workflows/CI/badge.svg)
+[![Gem Version](https://img.shields.io/gem/v/libssw?color=brightgreen)](https://rubygems.org/gems/libssw)
+[![Docs Latest](https://img.shields.io/badge/docs-latest-blue.svg)](https://rubydoc.info/gems/libssw)
 
 :checkered_flag: [libssw](https://github.com/mengyao/Complete-Striped-Smith-Waterman-Library) - fast SIMD parallelized implementation of the Smith-Waterman algorithm - for Ruby
 
@@ -35,18 +37,64 @@ bundle exec rake install
 ```ruby
 require 'libssw'
 
-ref = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
-read = [0, 1, 2, 3, 3]
-mat = [2, -2, -2, -2,  0,
-      -2,  2, -2, -2,  0,
-      -2, -2,  2, -2,  0,
-      -2, -2, -2,  2,  0,
-       0,  0,  0,  0,  0]
-profile = LibSSW.ssw_init(read, mat)
-align   = LibSSW.ssw_align(profile, ref, 3, 1, 1, 0, 0, 15)
-p align.to_h
+SSW = LibSSW
+
+ref_str  = "AAAAAAAAACGTTAAAAAAAAAA"
+ref_int  = SSW.dna_to_int_array(ref_str) 
+# [0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
+
+read_str1 = "ACGTT"
+read_str2 = SSW.dna_complement(read_str1)
+read_int1 = SSW.dna_to_int_array(read_str1)
+# [0, 1, 2, 3, 3]
+read_int2 = SSW.dna_to_int_array(read_str2)
+# [0, 0, 1, 2, 3]
+
+mat = SSW.create_scoring_matrix(SSW::DNAElements, 2, -2)
+# mat = [2, -2, -2, -2,  0,
+#       -2,  2, -2, -2,  0,
+#       -2, -2,  2, -2,  0,
+#       -2, -2, -2,  2,  0,
+#        0,  0,  0,  0,  0]
+
+profile1 = LibSSW.ssw_init(read_int1, mat)
+align1   = LibSSW.ssw_align(profile1, ref_int, 3, 1, 1, 0, 0, 15)
+pp align1.to_h
+# {
+#  :score1       => 10,
+#  :score2       => 0,
+#  :ref_begin1   => 8,
+#  :ref_end1     => 12,
+#  :read_begin1  => 0,
+#  :read_end1    => 4,
+#  :ref_end2     => 0,
+#  :cigar        => [80],
+#  :cigar_len    => 1,
+#  :cigar_string => "5M"
+# }
+
+profile2 = LibSSW.ssw_init(read_int2, mat)
+align2     = LibSSW.ssw_align(profile2, ref_int, 3, 1, 1, 0, 0, 15)
+pp align2.to_h
+# {
+#  :score1       => 10,
+#  :score2       => 0,
+#  :ref_begin1   => 7,
+#  :ref_end1     => 11,
+#  :read_begin1  => 0,
+#  :read_end1    => 4,
+#  :ref_end2     => 0,
+#  :cigar        => [80],
+#  :cigar_len    => 1,
+#  :cigar_string => "5M"
+# }
 ```
 
+
+## Documentation
+
+* [API Documentation](https://rubydoc.info/gems/libssw)
+
 ## Development
 
 ```sh

data/exe/rbssw

@@ -99,13 +99,9 @@ lScore = nil
 if opts[:bprotein]
   # load AA score matrix
   if !opts[:smatrix]
-    lEle = ['A', 'R', 'N', 'D', 'C', 'Q', 'E', 'G', 'H', 'I', 'L', 'K',
-            'M', 'F', 'P', 'S', 'T', 'W', 'Y', 'V', 'B', 'Z', 'X', '*']
-    lEle.each_with_index do |ele, i|
-      dEle2Int[ele] = i
-      dEle2Int[ele.downcase] = i
-      dInt2Ele[i] = ele
-    end
+    lEle = SSW::AAELEMENTS
+    dEle2Int = SSW::AA2INT
+    dInt2Ele = SSW::INT2AA
     nEleNum = lEle.size
     lScore = SSW::Blosum50
   else
@@ -113,31 +109,12 @@ if opts[:bprotein]
   end
 elsif !opts[:smatrix]
   # init DNA score matrix
-  lEle = %w[A C G T N]
-  dRc = { 'A': 'C', 'C': 'G', 'G': 'C', 'T': 'A', 'a': 'C', 'c': 'G', 'g': 'C', 't': 'A' }
-  lEle.each_with_index do |ele, i|
-    dEle2Int[ele] = i
-    dEle2Int[ele.downcase] = i
-    dInt2Ele[i] = ele
-  end
-  # dEle2Int = {'A': 0, 'a': 0, 'C': 1, 'G': 2, 'g': 2, 'c': 1, 'N': 4, 'T': 3, 'n': 4, 't': 3}
-  # dInt2Ele = {0: 'A', 1: 'C', 2: 'G', 3: 'T', 4: 'N'}
+  lEle = SSW::DNAELEMENTS
+  dRc = SSW::DNARC
+  dEle2Int = SSW::DNA2INT
+  dInt2Ele = SSW::INT2DNA
   nEleNum = lEle.size # 5
-  lScore = Array.new(nEleNum**2, 0)
-  (nEleNum - 1).times do |i|
-    (nEleNum - 1).times do |j|
-      lScore[i * nEleNum + j] = if lEle[i] == lEle[j]
-                                  opts[:nmatch]
-                                else
-                                  -opts[:nmismatch]
-                                end
-    end
-  end
-  # lScore = [ 2, -2, -2, -2, 0,
-  #           -2,  2, -2, -2, 0,
-  #           -2, -2,  2, -2, 0,
-  #           -2, -2, -2,  2, 0,
-  #            0,  0,  0,  0, 0 ]
+  lScore = LibSSW.create_scoring_matrix(lEle, opts[:nmatch], -opts[:nmismatch])
 end
 
 warn 'Reverse complement alignment is not available for protein sequences.' if opts[:bbest] && opts[:bprotein]

data/lib/libssw.rb

@@ -30,6 +30,64 @@ module LibSSW
   require_relative 'libssw/profile'
   require_relative 'libssw/align'
 
+  AAELEMENTS = ['A', 'R', 'N', 'D', 'C', 'Q', 'E', 'G',
+                'H', 'I', 'L', 'K', 'M', 'F', 'P', 'S',
+                'T', 'W', 'Y', 'V', 'B', 'Z', 'X', '*']
+
+  AA2INT = { 'A' => 0,  'a' => 0,
+             'R' => 1,  'r' => 1,
+             'N' => 2,  'n' => 2,
+             'D' => 3,  'd' => 3,
+             'C' => 4,  'c' => 4,
+             'Q' => 5,  'q' => 5,
+             'E' => 6,  'e' => 6,
+             'G' => 7,  'g' => 7,
+             'H' => 8,  'h' => 8,
+             'I' => 9,  'i' => 9,
+             'L' => 10, 'l' => 10,
+             'K' => 11, 'k' => 11,
+             'M' => 12, 'm' => 12,
+             'F' => 13, 'f' => 13,
+             'P' => 14, 'p' => 14,
+             'S' => 15, 's' => 15,
+             'T' => 16, 't' => 16,
+             'W' => 17, 'w' => 17,
+             'Y' => 18, 'y' => 18,
+             'V' => 19, 'v' => 19,
+             'B' => 20, 'b' => 20,
+             'Z' => 21, 'z' => 21,
+             'X' => 22, 'x' => 22,
+             '*' => 23 }
+
+  INT2AA = { 0 => 'A', 1 => 'R', 2 => 'N', 3 => 'D',
+             4 => 'C', 5 => 'Q', 6 => 'E', 7 => 'G',
+             8 => 'H', 9 => 'I', 10 => 'L', 11 => 'K',
+             12 => 'M', 13 => 'F', 14 => 'P', 15 => 'S',
+             16 => 'T', 17 => 'W', 18 => 'Y', 19 => 'V',
+             20 => 'B', 21 => 'Z', 22 => 'X', 23 => '*' }
+
+  DNAElements = %w[A C G T N]
+
+  DNA2INT = { 'A' => 0, 'a' => 0,
+              'C' => 1, 'c' => 1,
+              'G' => 2, 'g' => 2,
+              'T' => 3, 't' => 3,
+              'N' => 4, 'n' => 4 }
+
+  INT2DNA = { 0 => 'A', 1 => 'C', 2 => 'G', 3 => 'T', 4 => 'N' }
+
+  # reverse complement
+  DNARC = { 'A' => 'T',
+            'C' => 'G',
+            'G' => 'C',
+            'T' => 'A',
+            'N' => 'N',
+            'a' => 'T',
+            'c' => 'G',
+            'g' => 'C',
+            't' => 'A',
+            'n' => 'N' }
+
   class << self
     # Create the query profile using the query sequence.
     # @param read [Array] query sequence; the query sequence needs to be numbers
@@ -56,27 +114,19 @@ module LibSSW
         n,
         score_size
       )
-      profile = LibSSW::Profile.new(ptr)
-      # Check Garbage Collection
-      %i[read read_len mat n].zip([read, read_len, mat, n]).each do |name, obj|
-        next unless profile.public_send(name) != obj
-
-        warn "[Error] Struct member: '#{name}'"
-        warn "        * expected value: #{obj}"
-        warn "        *   actual value: #{profile.public_send(name)}"
-        warn "        This may have been caused by Ruby'S GC."
-      end
-      # Preventing Garbage Collection --force
-      cstruct         = profile.cstruct
-      cstruct.read    = read_str
-      cstruct.mat     = mat_str
-      cstruct.readLen = read_len
-      cstruct.n       = n
-      ptr.instance_variable_set(:@read_str, read_str)
-      ptr.instance_variable_set(:@read_len, read_len)
-      ptr.instance_variable_set(:@mat_str,  mat_str)
-      ptr.instance_variable_set(:@n,        n)
-      profile
+      # Garbage collection workaround
+      #
+      # * The following code will cause a segmentation violation when manually
+      #   releasing memory. The reason is unknown.
+      # * func_map is only available in newer versions of fiddle.
+      # ptr.free = FFI.instance_variable_get(:@func_map)['init_destroy']
+      ptr.instance_variable_set(:@read_str,   read_str)
+      ptr.instance_variable_set(:@read_len,   read_len)
+      ptr.instance_variable_set(:@mat_str,    mat_str)
+      ptr.instance_variable_set(:@n,          n)
+      ptr.instance_variable_set(:@score_size, score_size)
+
+      LibSSW::Profile.new(ptr)
     end
 
     # Release the memory allocated by function ssw_init.
@@ -140,6 +190,8 @@ module LibSSW
       # Not sure yet if we should set the instance variable to the pointer as a
       # garbage collection workaround.
       # For example: instance_variable_set(:@ref_str, ref_str)
+      #
+      # ptr.free = FFI.instance_variable_get(:@func_map)['align_destroy']
       LibSSW::Align.new(ptr)
     end
 
@@ -189,5 +241,61 @@ module LibSSW
       end
       cigar_string
     end
+
+    # Create scoring matrix of Smith-Waterman algrithum.
+    # @param [Array] elements
+    # @param [Integer] match_score
+    # @param [Integer] mismatch_score
+    def create_scoring_matrix(elements, match_score, mismatch_score)
+      size = elements.size
+      score = Array.new(size * size, 0)
+      (size - 1).times do |i|
+        (size - 1).times do |j|
+          score[i * size + j] = \
+            (elements[i] == elements[j] ? match_score : mismatch_score)
+        end
+      end
+      score
+    end
+
+    # @param [String] seq
+    def dna_to_int_array(seq)
+      raise ArgumentError, 'seq must be a string' unless seq.is_a? String
+
+      seq.each_char.map do |base|
+        DNA2INT[base] || DNA2INT['N']
+      end
+    end
+
+    def dna_complement(seq)
+      seq.each_char.map do |base|
+        DNARC[base]
+      end.join.reverse
+    end
+
+    # @param [Array] int array
+    def int_array_to_dna(arr)
+      raise ArgumentError, 'arr must be an Array' unless arr.is_a? Array
+
+      arr.map do |i|
+        INT2DNA[i] || 'N'
+      end.join
+    end
+
+    def aaseq_to_int_array(seq)
+      raise ArgumentError, 'seq must be a string' unless seq.is_a? String
+
+      seq.each_char.map do |base|
+        AA2INT[base] || AA2INT['*']
+      end
+    end
+
+    def int_array_to_aaseq(arr)
+      raise ArgumentError, 'arr must be an Array' unless arr.is_a? Array
+
+      arr.map do |i|
+        INT2AA[i] || '*'
+      end.join
+    end
   end
 end

data/lib/libssw/align.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative 'struct_helper'
-
 module LibSSW
   # structure of the alignment result
   # @!attribute score1
@@ -31,15 +29,13 @@ module LibSSW
   #   @return [Integer]
   #     length of the cigar string; cigarLen = 0 when the best alignment path is not available
   class Align < FFI::Align
-    include StructHelper
-
     def self.keys
       %i[score1 score2 ref_begin1 ref_end1
          read_begin1 read_end1 ref_end2 cigar cigar_len cigar_string]
     end
 
     # This class is read_only
-    attr_reader(*keys, :ptr, :cstruct)
+    attr_reader(*keys)
 
     def initialize(ptr)
       @ptr          = ptr
@@ -55,6 +51,11 @@ module LibSSW
       @cigar        = cigar_len.positive? ? align.cigar[0, 4 * cigar_len].unpack('L*') : []
       # Attributes for ruby binding only
       @cigar_string = LibSSW.array_to_cigar_string(@cigar)
+      LibSSW.align_destroy(ptr)
+    end
+
+    def to_h
+      self.class.keys.map { |k| [k, __send__(k)] }.to_h
     end
   end
 end

data/lib/libssw/profile.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative 'struct_helper'
-
 module LibSSW
   # structure of the query profile/usr/lib/x86_64-linux-gnu/
   # @!attribute read
@@ -10,8 +8,6 @@ module LibSSW
   # @!attribute n
   # @!attribute bias
   class Profile < FFI::Profile
-    include StructHelper
-
     def self.keys
       %i[read mat read_len n bias]
     end
@@ -31,11 +27,16 @@ module LibSSW
 
     def to_ptr
       # Garbage collection warkaround
-      # cstruct.read    = p @ptr.instance_variable_get(:@read_str)
-      # cstruct.mat     = p @ptr.instance_variable_get(:@mat_str)
-      # cstruct.readLen = p @ptr.instance_variable_get(:@read_len)
-      # cstruct.n       = p @ptr.instance_variable_get(:@n)
-      @ptr
+      # Preventing Garbage Collection --force
+      cstruct.read    = ptr.instance_variable_get(:@read_str)
+      cstruct.mat     = ptr.instance_variable_get(:@mat_str)
+      cstruct.readLen = ptr.instance_variable_get(:@read_len)
+      cstruct.n       = ptr.instance_variable_get(:@n)
+      ptr
+    end
+
+    def to_h
+      self.class.keys.map { |k| [k, __send__(k)] }.to_h
     end
   end
 end

data/lib/libssw/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module LibSSW
-  VERSION = '0.0.1'
+  VERSION = '0.0.2'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: libssw
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - kojix2
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2021-01-13 00:00:00.000000000 Z
+date: 2021-01-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: fiddle
@@ -125,7 +125,6 @@ files:
 - lib/libssw/align.rb
 - lib/libssw/ffi.rb
 - lib/libssw/profile.rb
-- lib/libssw/struct_helper.rb
 - lib/libssw/version.rb
 homepage: https://github.com/kojix2/ruby-libssw
 licenses:

data/lib/libssw/struct_helper.rb

@@ -1,13 +0,0 @@
-# frozen_string_literal: true
-
-module LibSSW
-  module StructHelper
-    def to_h
-      self.class.keys.map { |k| [k, __send__(k)] }.to_h
-    end
-
-    def to_ptr
-      @ptr
-    end
-  end
-end