bio-alignment 0.0.5 → 0.0.6

data/lib/bio-alignment/rows.rb

@@ -0,0 +1,52 @@
+require 'bio-alignment/state'
+
+module Bio
+ 
+  module BioAlignment
+
+    # The Rows module provides accessors for the Row list
+    # returning Row objects
+    module Rows
+     
+      # Return an copy of an alignment which matching rows. The originating
+      # sequences should have methods 'empty_copy' and '<<'
+      def rows_where &block
+        seqs = []
+        rows.each do | seq | 
+          seqs << seq.clone if block.call(seq)
+        end
+        Alignment.new(seqs)
+      end
+
+    end
+
+    # Support the notion of Rows in an alignment. A Row
+    # can have state by attaching state objects
+    class Row
+      include State
+
+      def initialize aln, row
+        @aln = aln
+        @row = row
+      end
+
+      def count &block
+        counter = 0
+        each do | e |
+          found = 
+            if e.kind_of?(String)
+              block.call(Element.new(e))
+            else
+              block.call(e)
+            end
+          counter += 1 if found
+        end
+        counter
+      end
+
+    end
+
+  end
+
+end
+

data/lib/bio-alignment/sequence.rb

@@ -1,20 +1,8 @@
 module Bio
   module BioAlignment
 
-    # Simple element that can be queried
-    class Element
-      def initialize c
-        @c = c
-      end
-      def gap?
-        @c == '-'
-      end
-      def to_s
-        @c
-      end
-    end
-
-    # A Sequence is a simple container for String sequences/lists
+    # A Sequence is a simple and efficient container for String sequences. To
+    # add state to elements unpack it into an Elements object with to_elements.
     #
     class Sequence
       include Enumerable
@@ -34,6 +22,8 @@ module Bio
         @seq.length
       end
 
+      # Return each element in the Sequence as an Element opbject, so it
+      # can be queried for gap? and undefined?
       def each
         @seq.each_char { | c | yield Element.new(c) }
       end
@@ -42,6 +32,22 @@ module Bio
         @seq.to_s
       end
 
+      def << element
+        @seq += element.to_s
+      end
+
+      def empty_copy
+        Sequence.new(@id,"")
+      end
+
+      def clone
+        Sequence.new(@id,@seq.clone)
+      end
+
+      # Return Sequence (string) as an Elements object
+      def to_elements
+        Elements.new(@id,@seq)
+      end
     end
   end
 end

data/lib/bio-alignment/state.rb

@@ -6,24 +6,80 @@ module Bio
       attr_accessor :state
     end
 
-    # Convenience class for tracking state. Note you can add
-    # any class you like
-    class ColumnState
+    module DeleteState
       attr_accessor :deleted
 
+      def delete!
+        @deleted = true
+      end
+
       def deleted?
-        deleted == true
+        @deleted == true
+      end
+
+      def to_s
+        (deleted? ? 'X' : '.')
+      end
+    end
+
+    module MarkState
+      attr_accessor :marked
+
+      def mark!
+        @marked = true
+      end
+
+      def unmark!
+        @marked = false
+      end
+
+      def marked?
+        @marked == true
+      end
+
+      def to_s
+        (marked? ? 'X' : '.')
+      end
+    end
+
+    module MaskState
+      attr_accessor :masked
+
+      def mask!
+        @masked = true
+      end
+
+      def unmask!
+        @masked = false
+      end
+
+      def masked?
+        @masked == true
+      end
+
+      def to_s
+        (masked? ? 'X' : '.')
       end
     end
 
+    # Convenience class for tracking state. Note you can add
+    # any class you like
+    class ColumnState
+      include DeleteState
+    end
+
     # Convenience class for tracking state. Note you can add
     # any class you like
     class RowState
-      attr_accessor :deleted
+      include DeleteState
+    end
 
-      def deleted?
-        deleted == true
-      end
+    class ElementState
+      include MarkState
+    end
+
+    class ElementMaskedState
+      include MaskState
     end
 
   end

data/lib/bio-alignment/tree.rb

@@ -0,0 +1,77 @@
+module Bio
+
+  module BioAlignment
+
+    # The Tree module turns a tree into a traversable object, by wrapping
+    # BioRuby's basic tree objects. The Bio::Tree object can always be
+    # fetched using to_bioruby_tree.
+
+    module Tree
+
+      class Node
+      end
+
+      # Make all nodes in the Bio::Tree aware of the tree object so we can use
+      # its methods
+      def Tree::init tree
+        if tree.kind_of?(Bio::Tree)
+          # walk all nodes and infect the tree info
+          tree.each_node do | node |
+            node.inject_tree(tree)
+          end
+          # tree.root.set_tree(tree)
+        else
+          raise "BioAlignment::Tree does not understand tree type "+tree.class.to_s
+        end
+        return tree
+      end
+
+      def root
+        @tree.root
+      end
+
+    end
+
+  end
+
+  # Here we add to BioRuby's Bio::Tree classes
+  class Tree
+    class Node
+      def inject_tree tree
+        @tree = tree
+      end
+
+      def leaf?
+        children.size == 0
+      end
+
+      def children
+        @tree.children(self)
+      end
+
+      def parent
+        @tree.parent(self)
+      end
+  
+      # Get the distance to another node (FIXME: write test)
+      def distance other
+        @tree.distance(self,other)
+      end
+    end
+
+    def find name
+      get_node_by_name(name)
+    end
+
+    # Walk the ordered tree leaves, calling into the block, and return an array 
+    def map 
+      res = []
+      leaves.each do | leaf |
+        item = yield leaf
+        res << item
+      end
+      res
+    end
+
+  end
+end

data/spec/bio-alignment_spec.rb

@@ -1,6 +1,7 @@
 require File.expand_path(File.dirname(__FILE__) + '/spec_helper')
 
-require 'bigbio'
+require 'bio-alignment'
+require 'bigbio'  # for the FastaReader
 include Bio::BioAlignment # Namespace
 
 describe "BioAlignment::CodonSequence" do
@@ -49,6 +50,61 @@ describe "BioAlignment::Alignment" do
     aln.sequences.first.seq[0].to_aa.should == "M"
     aln.sequences.first.seq[2].to_aa.should == "T"
   end
+end
+
 
+describe "BioAlignment::DelBridges" do
+  require 'bio-alignment/edit/del_bridges'
+    string = 
+      """
+      ----SNSFSRPTIIFSGCSTACSGK--SELVCGFRSFMLSDV
+      SSIISNSFSRPTIIFSGCSTACSGK--SEQVCGFR---LSDV
+      SSIISNSFSRPTIIFSGCSTACSGKLTSEQVCGFR---LSDV
+      ----PKLFSRPTIIFSGCSTACSGK--SEPVCGFRSFMLSDV
+      ----------PTIIFSGCSKACSGKGLSELVCGFRSFMLSDV
+      ----------PTIIFSGCSKACSGK-----FRSFRSFMLSAV
+      ----------PTIIFSGCSKACSGK-----VCGIFHAVRSFM
+      ----------PTIIFSGCSKACSGK--SELVCGFRSFMLSAV
+      -------------IFHAVR-TC-HP-----------------
+      """
+  aln = Alignment.new(string.split(/\n/))
+  print aln.to_s,"\n"
+  columns = aln.columns
+  columns.should_not == nil
+  columns.should_not == []
+  columns.size.should == 42
+  # make sure we are using the same columns
+  aln.columns.should == columns
+  aln.extend DelBridges
+  aln2 = aln.mark_bridges
+  print aln2.to_s,"\n"
+  columns2 = aln2.columns
+  columns2.should_not == nil
+  columns2.should_not == [] 
+  columns2.count { |col| col.state.deleted? }.should == 6
+  aln2.columns[0].state.should == columns2[0].state
+  aln2.columns[0].state.should_not == columns[0].state
+  aln2.rows.first.to_s.should == "----SNSFSRPTIIFSGCSTACSGK--SELVCGFRSFMLSDV"
+  # now write out the alignment with deleted columns removed
+  aln3 = aln2.columns_where { |col| !col.state.deleted? }
+  print aln3.to_s,"\n"
+  aln3.rows.first.to_s.should == "SNSFSRPTIIFSGCSTACSGKSELVCGFRSFMLSDV"
 end
 
+describe "BioAlignment::DelBridges for codons" do
+  # We are going to do the same for a codon alignment
+  aln = Alignment.new
+  fasta = FastaReader.new('test/data/fasta/codon/codon-alignment.fa')
+  fasta.each do | rec |
+    aln.sequences << CodonSequence.new(rec.id, rec.seq)
+  end
+  aln.extend DelBridges
+  aln2 = aln.mark_bridges
+  # print aln2[0].to_s,"\n"
+  aln2.columns.size.should == 404
+  # count deleted columns
+  aln2.columns.count { |col| col.state.deleted? }.should == 5
+  # create new alignment
+  aln3 = aln2.columns_where { |col| !col.state.deleted? }
+  aln3.columns.size.should == 399
+end

data/spec/spec_helper.rb

@@ -7,6 +7,6 @@ require 'bio-alignment'
 # in ./support/ and its subdirectories.
 Dir["#{File.dirname(__FILE__)}/support/**/*.rb"].each {|f| require f}
 
-RSpec.configure do |config|
-  
-end
+# RSpec.configure do |config|
+#   
+# end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: bio-alignment
 version: !ruby/object:Gem::Version
-  version: 0.0.5
+  version: 0.0.6
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-02-28 00:00:00.000000000Z
+date: 2012-03-17 00:00:00.000000000Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bio-logger
-  requirement: &11633860 !ruby/object:Gem::Requirement
+  requirement: &26202820 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -21,10 +21,10 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *11633860
+  version_requirements: *26202820
 - !ruby/object:Gem::Dependency
   name: bio
-  requirement: &11632600 !ruby/object:Gem::Requirement
+  requirement: &26201340 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -32,10 +32,21 @@ dependencies:
         version: 1.4.2
   type: :runtime
   prerelease: false
-  version_requirements: *11632600
+  version_requirements: *26201340
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: &26199400 !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: *26199400
 - !ruby/object:Gem::Dependency
   name: bio-bigbio
-  requirement: &11622800 !ruby/object:Gem::Requirement
+  requirement: &26197880 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>'
@@ -43,10 +54,10 @@ dependencies:
         version: 0.1.3
   type: :development
   prerelease: false
-  version_requirements: *11622800
+  version_requirements: *26197880
 - !ruby/object:Gem::Dependency
   name: cucumber
-  requirement: &11622000 !ruby/object:Gem::Requirement
+  requirement: &26196760 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -54,10 +65,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *11622000
+  version_requirements: *26196760
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &11621400 !ruby/object:Gem::Requirement
+  requirement: &26195120 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -65,29 +76,29 @@ dependencies:
         version: 2.3.0
   type: :development
   prerelease: false
-  version_requirements: *11621400
+  version_requirements: *26195120
 - !ruby/object:Gem::Dependency
   name: bundler
-  requirement: &11620480 !ruby/object:Gem::Requirement
+  requirement: &26194620 !ruby/object:Gem::Requirement
     none: false
     requirements:
-    - - ~>
+    - - ! '>='
       - !ruby/object:Gem::Version
-        version: 1.0.0
+        version: 1.0.21
   type: :development
   prerelease: false
-  version_requirements: *11620480
+  version_requirements: *26194620
 - !ruby/object:Gem::Dependency
   name: jeweler
-  requirement: &11619680 !ruby/object:Gem::Requirement
+  requirement: &26193920 !ruby/object:Gem::Requirement
     none: false
     requirements:
-    - - ~>
+    - - ! '>='
       - !ruby/object:Gem::Version
-        version: 1.7.0
+        version: '0'
   type: :development
   prerelease: false
-  version_requirements: *11619680
+  version_requirements: *26193920
 description: Alignment handler for multiple sequence alignments (MSA)
 email: pjotr.public01@thebird.nl
 executables:
@@ -115,7 +126,10 @@ files:
 - features/columns.feature
 - features/edit/del_bridges-feature.rb
 - features/edit/del_bridges.feature
+- features/edit/del_non_informative_sequences-feature.rb
 - features/edit/del_non_informative_sequences.feature
+- features/edit/del_short_sequences-feature.rb
+- features/edit/del_short_sequences.feature
 - features/edit/gblocks-feature.rb
 - features/edit/gblocks.feature
 - features/edit/mask_islands-feature.rb
@@ -126,15 +140,26 @@ files:
 - features/pal2nal.feature
 - features/rows-feature.rb
 - features/rows.feature
+- features/tree-feature.rb
+- features/tree.feature
 - lib/bio-alignment.rb
 - lib/bio-alignment/alignment.rb
 - lib/bio-alignment/bioruby.rb
 - lib/bio-alignment/codonsequence.rb
-- lib/bio-alignment/column.rb
+- lib/bio-alignment/columns.rb
 - lib/bio-alignment/edit/del_bridges.rb
+- lib/bio-alignment/edit/del_non_informative_sequences.rb
+- lib/bio-alignment/edit/del_short_sequences.rb
+- lib/bio-alignment/edit/edit_columns.rb
+- lib/bio-alignment/edit/edit_rows.rb
+- lib/bio-alignment/edit/mask_islands.rb
+- lib/bio-alignment/edit/mask_serial_mutations.rb
+- lib/bio-alignment/elements.rb
 - lib/bio-alignment/pal2nal.rb
+- lib/bio-alignment/rows.rb
 - lib/bio-alignment/sequence.rb
 - lib/bio-alignment/state.rb
+- lib/bio-alignment/tree.rb
 - spec/bio-alignment_spec.rb
 - spec/spec_helper.rb
 - test/data/fasta/codon/aa-alignment.fa
@@ -158,7 +183,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: 1565072942973090495
+      hash: 1800672102634743595
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements: