lederhosen 1.1.1 → 1.2.0

data/lederhosen.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = "lederhosen"
-  s.version = "1.1.1"
+  s.version = "1.2.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Austin G. Davis-Richardson"]
-  s.date = "2012-10-30"
+  s.date = "2012-11-13"
   s.description = "Various tools for OTU clustering"
   s.email = "harekrishna@gmail.com"
   s.executables = ["lederhosen"]
@@ -26,7 +26,9 @@ Gem::Specification.new do |s|
     "lederhosen.gemspec",
     "lib/lederhosen.rb",
     "lib/lederhosen/cli.rb",
+    "lib/lederhosen/no_tasks.rb",
     "lib/lederhosen/tasks/cluster.rb",
+    "lib/lederhosen/tasks/get_reps.rb",
     "lib/lederhosen/tasks/join_otu_tables.rb",
     "lib/lederhosen/tasks/make_udb.rb",
     "lib/lederhosen/tasks/otu_filter.rb",
@@ -48,7 +50,7 @@ Gem::Specification.new do |s|
   s.homepage = "http://audy.github.com/lederhosen"
   s.licenses = ["MIT"]
   s.require_paths = ["lib"]
-  s.rubygems_version = "1.8.24"
+  s.rubygems_version = "1.8.23"
   s.summary = "OTU Clustering"
 
   if s.respond_to? :specification_version then

data/lib/lederhosen/no_tasks.rb

@@ -0,0 +1,56 @@
+module Lederhosen
+  class CLI
+    no_tasks do
+
+      # parse a line of usearch prefix
+      # return a hash in the form:
+      # { :taxonomy => '', :identity => 0.00, ... }
+      # unless the line is not a "hit" in which case
+      # the function returns nil
+      def parse_usearch_line(str)
+
+        # skip non hits
+        return nil unless str =~ /^H/
+
+        str = str.split
+
+        taxonomic_description = str[9]
+        identity = str[3].to_f
+
+        # parse taxonomic_description
+        taxonomies = parse_taxonomy(taxonomic_description)
+
+        { :identity => identity }.merge(taxonomies)
+      end
+
+      # parse a taxonomic description using the
+      # taxcollector format returning name at each level (genus, etc...)
+      def parse_taxonomy(taxonomy)
+
+        levels = { 'domain'  => 0,
+                   'kingdom' => 0,
+                   'phylum'  => 1,
+                   'class'   => 2,
+                   'order'   => 3,
+                   'family'  => 4,
+                   'genus'   => 5,
+                   'species' => 6 }
+
+        names = Hash.new
+
+        levels.each_pair do |level, num|
+          name = taxonomy.match(/\[#{num}\](\w*)[;\[]/)[1] rescue nil
+          names[level] = name
+        end
+
+        # keep original taxonomic description
+        names[:original] = taxonomy
+
+        names
+      end
+
+    end # no tasks
+
+  end
+end
+

data/lib/lederhosen/tasks/get_reps.rb

@@ -0,0 +1,47 @@
+require 'set'
+
+module Lederhosen
+  class CLI
+    desc 'get_reps', 'get representative reads for a uc file'
+
+    method_option :input,    :type => :string, :required => true
+    method_option :database, :type => :string, :required => true
+    method_option :output,   :type => :string, :required => true
+
+    def get_reps
+      input    = options[:input]
+      database = options[:database]
+      output   = options[:output]
+
+      taxa = Set.new
+
+      ohai "getting representative database sequences from #{database} using #{input} clusters and saving to #{output}"
+
+      # parse uc file, get list of taxa we need to get
+      # full sequences for from the database
+      File.open(input).each do |line|
+        header = parse_usearch_line(line.strip)
+        taxa << header[:original] rescue nil
+      end
+
+      ohai "found #{taxa.size} representative sequences"
+
+      # print representative sequences from database
+      output = File.open(output, 'w')
+      kept = 0
+      File.open(database) do |handle|
+        Dna.new(handle).each do |record|
+          if taxa.include? record.name
+            output.puts record
+            kept += 1
+          end
+        end
+      end
+
+      output.close
+
+      ohai "saved #{kept} representatives"
+
+    end
+  end
+end

data/lib/lederhosen/tasks/otu_table.rb

@@ -81,52 +81,6 @@ module Lederhosen
       end
     end
 
-    no_tasks do
-      # parse a line of usearch prefix
-      # return a hash in the form:
-      # { :taxonomy => '', :identity => 0.00, ... }
-      # unless the line is not a "hit" in which case
-      # the function returns nil
-      def parse_usearch_line(str)
-
-        # skip non hits
-        return nil unless str =~ /^H/
-
-        str = str.split
-
-        taxonomic_description = str[9]
-        identity = str[3].to_f
-
-        # parse taxonomic_description
-        taxonomies = parse_taxonomy(taxonomic_description)
-
-        { :identity => identity }.merge(taxonomies)
-      end
-
-      # parse a taxonomic description using the
-      # taxcollector format returning name at each level (genus, etc...)
-      def parse_taxonomy(taxonomy)
-
-        levels = { 'domain'  => 0,
-                   'kingdom' => 0,
-                   'phylum'  => 1,
-                   'class'   => 2,
-                   'order'   => 3,
-                   'family'  => 4,
-                   'genus'   => 5,
-                   'species' => 6 }
-
-        names = Hash.new
-
-        levels.each_pair do |level, num|
-          name = taxonomy.match(/\[#{num}\](\w*)[;\[]/)[1] rescue nil
-          names[level] = name
-        end
-
-        names
-      end
-
-    end # no tasks
 
   end # class CLI
 end # module Lederhosen

data/lib/lederhosen/version.rb

@@ -1,9 +1,9 @@
 module Lederhosen
   module Version
     MAJOR = 1
-    MINOR = 1
-    CODENAME = 'Apfelstrudel' # changes for minor versions
-    PATCH = 1
+    MINOR = 2
+    CODENAME = 'Regenmantel' # changes for minor versions
+    PATCH = 0
 
     STRING = [MAJOR, MINOR, PATCH].join('.')
   end

data/readme.md

@@ -1,3 +1,5 @@
+<img src="http://d.pr/i/26Js+#.png" align="right">
+
 # Lederhosen
 
 Cluster raw Illumina 16S rRNA amplicon data to generate OTUs.
@@ -46,19 +48,33 @@ The trimming process will reverse complement the "right" pair so that both reads
 
 Create UDB database required by usearch from TaxCollector
 
-    lederhosen make_udb --input=taxcollector.fa --output=taxcollector.udb
+```bash
+lederhosen make_udb \
+	--input=taxcollector.fa \
+	--output=taxcollector.udb
+```
 
 ### Cluster Reads using USEARCH
 
 Cluster reads using USEARCH. Output is a uc file.
 
-    lederhosen cluster --input=trimmed/*.fasta --identity=0.95 --output=clusters_95.uc --database=taxcollector.udb
-
+```bash
+lederhosen cluster \
+	--input=trimmed/*.fasta \
+	--identity=0.95 \
+	--output=clusters_95.uc \
+	--database=taxcollector.udb
+```
 ### Generate OTU table(s)
 
 Create an OTU abundance table where rows are samples and columns are clusters. The entries are the number of reads for that cluster in a sample.
 
-    lederhosen otu_table --clusters=clusters_95.uc --prefix=otu_table --level=domain phylum class order family genus species
+```bash
+lederhosen otu_table \
+	--clusters=clusters_95.uc \
+	--prefix=otu_table \
+	--level=domain phylum class order family genus species
+```
 
 This will create the files:
 

data/spec/cli_spec.rb

@@ -24,6 +24,8 @@ describe Lederhosen::CLI do
 
   it 'can cluster reads using usearch' do
     `./bin/lederhosen cluster --input #{$test_dir}/trimmed/ILT_L_9_B_001.fasta --database #{$test_dir}/test_db.udb --identity 0.95 --output #{$test_dir}/clusters.uc`
+    $?.success?.should be_true
+    File.exists?(File.join($test_dir, 'clusters.uc')).should be_true
   end
 
   it 'should build abundance matrices for each level' do
@@ -47,5 +49,9 @@ describe Lederhosen::CLI do
     $?.success?.should be_true
   end
 
+  it 'should print representative sequences from uc files' do
+    `./bin/lederhosen get_reps --input=#{$test_dir}/clusters.uc --database=#{$test_dir}/trimmed/ILT_L_9_B_001.fasta --output=#{$test_dir}/representatives.fasta`
+  end
+
   it 'should create a fasta file containing representative reads for each cluster'
 end

metadata

@@ -1,118 +1,120 @@
---- !ruby/object:Gem::Specification 
+--- !ruby/object:Gem::Specification
 name: lederhosen
-version: !ruby/object:Gem::Version 
-  hash: 17
+version: !ruby/object:Gem::Version
+  version: 1.2.0
   prerelease: 
-  segments: 
-  - 1
-  - 1
-  - 1
-  version: 1.1.1
 platform: ruby
-authors: 
+authors:
 - Austin G. Davis-Richardson
 autorequire: 
 bindir: bin
 cert_chain: []
-
-date: 2012-10-30 00:00:00 Z
-dependencies: 
-- !ruby/object:Gem::Dependency 
-  type: :runtime
-  requirement: &id001 !ruby/object:Gem::Requirement 
+date: 2012-11-13 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: dna
+  requirement: !ruby/object:Gem::Requirement
     none: false
-    requirements: 
-    - - "="
-      - !ruby/object:Gem::Version 
-        hash: 7
-        segments: 
-        - 0
-        - 0
-        - 12
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
         version: 0.0.12
-  version_requirements: *id001
-  name: dna
-  prerelease: false
-- !ruby/object:Gem::Dependency 
   type: :runtime
-  requirement: &id002 !ruby/object:Gem::Requirement 
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
     none: false
-    requirements: 
-    - - ">="
-      - !ruby/object:Gem::Version 
-        hash: 3
-        segments: 
-        - 0
-        version: "0"
-  version_requirements: *id002
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 0.0.12
+- !ruby/object:Gem::Dependency
   name: progressbar
-  prerelease: false
-- !ruby/object:Gem::Dependency 
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
   type: :runtime
-  requirement: &id003 !ruby/object:Gem::Requirement 
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
     none: false
-    requirements: 
-    - - ">="
-      - !ruby/object:Gem::Version 
-        hash: 3
-        segments: 
-        - 0
-        version: "0"
-  version_requirements: *id003
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
   name: thor
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
   prerelease: false
-- !ruby/object:Gem::Dependency 
-  type: :development
-  requirement: &id004 !ruby/object:Gem::Requirement 
+  version_requirements: !ruby/object:Gem::Requirement
     none: false
-    requirements: 
-    - - ~>
-      - !ruby/object:Gem::Version 
-        hash: 31
-        segments: 
-        - 3
-        - 12
-        version: "3.12"
-  version_requirements: *id004
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
   name: rdoc
-  prerelease: false
-- !ruby/object:Gem::Dependency 
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '3.12'
   type: :development
-  requirement: &id005 !ruby/object:Gem::Requirement 
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
     none: false
-    requirements: 
-    - - ">="
-      - !ruby/object:Gem::Version 
-        hash: 3
-        segments: 
-        - 0
-        version: "0"
-  version_requirements: *id005
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '3.12'
+- !ruby/object:Gem::Dependency
   name: bundler
-  prerelease: false
-- !ruby/object:Gem::Dependency 
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
   type: :development
-  requirement: &id006 !ruby/object:Gem::Requirement 
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
     none: false
-    requirements: 
-    - - ">="
-      - !ruby/object:Gem::Version 
-        hash: 3
-        segments: 
-        - 0
-        version: "0"
-  version_requirements: *id006
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
   name: jeweler
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
   prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
 description: Various tools for OTU clustering
 email: harekrishna@gmail.com
-executables: 
+executables:
 - lederhosen
 extensions: []
-
-extra_rdoc_files: 
+extra_rdoc_files:
 - LICENSE.txt
-files: 
+files:
 - .rspec
 - .rvmrc
 - Gemfile
@@ -122,7 +124,9 @@ files:
 - lederhosen.gemspec
 - lib/lederhosen.rb
 - lib/lederhosen/cli.rb
+- lib/lederhosen/no_tasks.rb
 - lib/lederhosen/tasks/cluster.rb
+- lib/lederhosen/tasks/get_reps.rb
 - lib/lederhosen/tasks/join_otu_tables.rb
 - lib/lederhosen/tasks/make_udb.rb
 - lib/lederhosen/tasks/otu_filter.rb
@@ -141,37 +145,31 @@ files:
 - spec/misc_spec.rb
 - spec/spec_helper.rb
 homepage: http://audy.github.com/lederhosen
-licenses: 
+licenses:
 - MIT
 post_install_message: 
 rdoc_options: []
-
-require_paths: 
+require_paths:
 - lib
-required_ruby_version: !ruby/object:Gem::Requirement 
+required_ruby_version: !ruby/object:Gem::Requirement
   none: false
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      hash: 3
-      segments: 
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+      segments:
       - 0
-      version: "0"
-required_rubygems_version: !ruby/object:Gem::Requirement 
+      hash: -2175708922821992201
+required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      hash: 3
-      segments: 
-      - 0
-      version: "0"
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
 requirements: []
-
 rubyforge_project: 
-rubygems_version: 1.8.24
+rubygems_version: 1.8.23
 signing_key: 
 specification_version: 3
 summary: OTU Clustering
 test_files: []
-