parse_fasta 1.7.2 → 1.8.0

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    NzY3MWVlYjVmYzM4NzhjZjIzNDUyYjFlNWZhMjMxMGE4NWQ3YTI4Yw==
+    MmIxNzgzM2E2MzkzYTc1ZjM2MWE0NzIxN2ZmZmRkMGUxMDQ0Y2MzYw==
   data.tar.gz: !binary |-
-    OTVjMThlYzg0NzUwZTdmMzIyZmJlYjc5NjBiNDUzMmE2NjJjMGVkOA==
+    Njc4ZTVlN2EwMTVjZDkwNWZmYTM2MzcwY2FkY2NlNjVmNmJiNWFmYw==
 SHA512:
   metadata.gz: !binary |-
-    ZTVkOTQ4MGMzMTQ4ZDg5MzczMDllMzgwNDY3Y2NmNzU4NDc5MzMwOGIwYzA3
-    Y2FjMzQ1YzA4NDhiMjY0OWMyYjMyMWFhOWM5MTI5OGFiMTJiMDFhYmJlNTI3
-    NjBkZDU5MzMzOTJlOGY0NzdiZWRiYWI5ZWE2YTBiZjJkYTk1ZTU=
+    ZjNiOTc2NTM0OTc0M2U4NDNkYzhlNTBlNjM3MzI4Mjc4YWY4YmE4MjRhMDdj
+    MGRiNWNiMmExZTZjYmRhZDk0YzcxMTQ1OGQ0NzdkYTVmMmQ0Mzg1YmU0NmFl
+    NmNjYWM1MmY4ZWRhNTY5MDAzNTk1YzcyN2IzYWE2MTNkMzAwNDg=
   data.tar.gz: !binary |-
-    MWJkZjY5ZWY5MmYyNTJjMjIyODE0ZmQ4YWU4OTZlMTY5MzIyNGNlZGJkNjU2
-    YjBjNmI5YTIyYTdmMGJhZmVlNTExZWJkN2U4OWRlMDE0NzM3MGI4ZGE4M2Rj
-    MzNjODU3YzYzMDFmZDg1MjU1YzQ2MDgxODIwNmQ3ZDJkZDU3MWQ=
+    Mjk5ZjU3YWI3YTJlN2Q4NWJjMWY3NDczOTBhNzI3NzlkMDViZjRhZGFkN2I4
+    NmU5YzFhYTI4ZDc2N2RhYTE5ODdkODE5NTQ5ZjJmNzNmNjEyNzY3NzJiZTk3
+    ODA2MWI3YjUzZDdmMTE5MGM1MDA3ZTk1NmMyNGU3NjFmOTIyMWY=

data/README.md

@@ -27,7 +27,7 @@ lightweight than BioRuby. And more fun! ;)
 ## Documentation ##
 
 Checkout
-[parse_fasta docs](http://rubydoc.info/gems/parse_fasta/1.7.2/frames)
+[parse_fasta docs](http://rubydoc.info/gems/parse_fasta/1.8.0/frames)
 for the full api documentation.
 
 ## Usage ##
@@ -66,6 +66,13 @@ Read fasta file into a hash.
 
 ## Versions ##
 
+### 1.8 ###
+
+Add `Sequence#rev_comp`. It can handle IUPAC characters. Since
+`parse_fasta` doesn't check whether the seq is AA or NA, if called on
+an amino acid string, things will get weird as it will complement the
+IUPAC characters in the AA string and leave others.
+
 ### 1.7 ###
 
 Add `SeqFile#to_hash`, `FastaFile#to_hash` and `FastqFile#to_hash`.

data/lib/parse_fasta/sequence.rb

@@ -20,6 +20,15 @@
 # nucleotide sequences.
 class Sequence < String
 
+  # # Error raised if both T and U are present
+  # #
+  # # @note This is NOT checked on every call to Sequence.new
+  # class AmbiguousSequenceError < StandardError
+  #   def message
+  #     "Sequence is ambiguous -- both T and U present"
+  #   end
+  # end
+
   # Strips whitespace from the str argument before calling super
   #
   # @return [Sequence] A Sequence string
@@ -130,4 +139,30 @@ class Sequence < String
       base_counts.map { |base, count| [base, count/total_bases] }.flatten
     Hash[*base_freqs]
   end
+
+  # Returns a reverse complement of self
+  #
+  # @return [Sequence] a Sequence that is the reverse complement of
+  #   self
+  #
+  # @example Hanldes any IUPAC character and capitalization properly
+  #   Sequence.new("gARKbdctymvhu").rev_comp #=> "adbkraghvMYTc"
+  #
+  # @example Leaves non IUPAC characters
+  #   Sequence.new("cccc--CCCcccga").rev_comp #=> "tcgggGGG--gggg""
+  #
+  # @note If Sequence contains non-IUPAC characters, these are not
+  #   complemented
+  def rev_comp
+    # if self.match(/T/i) && self.match(/U/i)
+    #   raise Sequence::AmbiguousSequenceError
+    # end
+
+    # if self.match(/[^ATUGCYRSWKMBDHVN]/i)
+    #   warn "WARNING: Sequence contains non IUPAC characters"
+    # end
+
+    self.reverse.tr("ATUGCYRSWKMBDHVNatugcyrswkmbdhvn",
+                    "TAACGRYSWMKVHDBNtaacgryswmkvhdbn")
+  end
 end

data/lib/parse_fasta/version.rb

@@ -17,5 +17,5 @@
 # along with parse_fasta.  If not, see <http://www.gnu.org/licenses/>.
 
 module ParseFasta
-  VERSION = "1.7.2"
+  VERSION = "1.8.0"
 end

data/spec/lib/sequence_spec.rb

@@ -21,6 +21,10 @@ require 'bio'
 
 describe Sequence do
 
+  # it "has AmbiguousSequenceError" do
+  #   expect(Sequence::AmbiguousSequenceError).not_to be nil
+  # end
+
   it "inherits from String" do
     expect(Sequence.new('ACTG')).to be_a String
   end
@@ -128,4 +132,49 @@ describe Sequence do
       end
     end
   end
+
+  describe "#rev_comp" do
+    # it "raises error if both T and U are present" do
+    #   s = Sequence.new("actGU")
+    #   err = Sequence::AmbiguousSequenceError
+    #   msg = "Sequence is ambiguous -- both T and U present"
+    #   expect { s.rev_comp }.to raise_error(err, msg)
+    # end
+
+    # it "warns if non iupac characters are present" do
+    #   s = Sequence.new("--..9284ldkjfalsjf")
+    #   msg = "WARNING: Sequence contains non IUPAC characters"
+    #   expect(s).to receive(:warn).with(msg)
+    #   s.rev_comp
+    # end
+    it "returns a reverse complement of the Sequence" do
+      s = Sequence.new("gARKbdctymvhu").rev_comp
+      expect(s).to eq "adbkraghvMYTc"
+
+      s = Sequence.new("ctyMVhgarKBda").rev_comp
+      expect(s).to eq "thVMytcdBKrag"
+    end
+
+    it "leaves non-IUPAC characters alone" do
+      s = Sequence.new("cccc--CCCcccga").rev_comp
+      expect(s).to eq "tcgggGGG--gggg"
+    end
+
+    it "returns a Sequence" do
+      s = Sequence.new("cccc--CCCcccga")
+      expect(s.rev_comp).to be_an_instance_of(Sequence)
+    end
+
+    it "gives back original sequence when called in succession" do
+      s = Sequence.new("cccc--CCCcccga")
+      expect(s.rev_comp.rev_comp).to eq s
+    end
+
+    context "with an empty sequence" do
+      it "returns an empty sequence" do
+        s = Sequence.new("")
+        expect(s.rev_comp).to be_empty
+      end
+    end
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: parse_fasta
 version: !ruby/object:Gem::Version
-  version: 1.7.2
+  version: 1.8.0
 platform: ruby
 authors:
 - Ryan Moore
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-10-19 00:00:00.000000000 Z
+date: 2015-11-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler