bio-ipcress 0.0.1 → 0.1.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 04f49808a34b7c50343255fc72f11eda8ca54de7
+  data.tar.gz: a9e7a50b28bba9ad139c065349291353d1ace24b
+SHA512:
+  metadata.gz: 5d29e0fef85c45118303a32469d305911815615300c018d3dd12c48537301bb92afeb4a6bb685aad0c6ddf522f2bb18efbec0e144502f49be22b5b8cc7d3a0b8
+  data.tar.gz: bf90fc0954bd8ba9f187f65d0c7d31676be8fa6abbf5d792b029745a17030c8e0520d54b05229180c058932dfaa03669efa3c9e3b20c8425d5c7d70e08837017

data/Gemfile

@@ -9,8 +9,7 @@ gem 'bio', '>=1.4.2'
 group :development do
   gem "shoulda", ">= 0"
   gem "rdoc", "~> 3.12"
-  gem "jeweler", "~> 1.8.3"
+  gem "jeweler", "~> 2.3", ">=2.3.2"
   gem "bundler", ">= 1.0.21"
-  gem "bio", ">= 1.4.2"
-  gem "rdoc", "~> 3.12"
+  gem 'test-unit'
 end

data/README.md

@@ -17,7 +17,7 @@ results = Bio::Ipcress.run(
   primer_set,
   'Methanocella_conradii_16s.fa', #this file is in the test/data/Ipcress directory
   {:min_distance => 2, :max_distance => 10000})
-#=> A Bio::Ipcress::Result object, parsed from
+#=> An array of Bio::Ipcress::Result objects, parsed from
 #
 #Ipcress result
 #--------------
@@ -59,8 +59,9 @@ res.forward_mismatches #=> 1
 res.reverse_mismatches #=> 1
 ```
 
-There appears to be a slight bug in iPCRess, in the way it handles primers with 'wobble' bases like AAACT*Y*,
-which indicates that both AAACT*C* and AAACT*T* are added as primers.
+There appears to be a slight bug in iPCRess, in the way it handles primers with 'wobble' bases like the
+last base of AAACTY,
+which indicates that both AAACTC and AAACTT are added as primers.
 IPCress always suggests that there is at least a single mismatch,
 when this is not always the case. To workaround this, the
 ```Result#recalculate_mismatches_from_alignments``` method re-computes the 

data/VERSION

@@ -1 +1 @@
-0.0.1
+0.1.0

data/bin/pcr.rb

@@ -0,0 +1,135 @@
+#!/usr/bin/env ruby
+
+require 'optparse'
+require 'bio-logger'
+require 'bio-ipcress'
+require 'bio'
+
+SCRIPT_NAME = File.basename(__FILE__); LOG_NAME = SCRIPT_NAME.gsub('.rb','')
+
+# Parse command line options into the options hash
+options = {
+  :logger => 'stderr',
+  :log_level => 'info',
+  :print_amplicon => false,
+  :num_mismatches => 3,
+}
+o = OptionParser.new do |opts|
+  opts.banner = "
+    Usage: #{SCRIPT_NAME} <arguments>
+
+    Take a set of primers and a fasta file of genome(s), and output each place that the primer hits \n\n"
+
+  opts.on("--primer1 PRIMER", "sequence of the forward primer [required]") do |arg|
+    options[:primer1] = arg
+  end
+  opts.on("--primer2 PRIMER", "sequence of the reverse primer [required]") do |arg|
+    options[:primer2] = arg
+  end
+  opts.on("--fasta FASTA_FILE[, FASTA_FILE2, ..]", Array, "sequence(s) being assayed [required]") do |arg|
+    options[:fasta] = arg
+  end
+  opts.on("--mismatches NUM", Integer, "max number of allowed mismatches. [default: #{options[:num_mismatches] }]") do |arg|
+    options[:num_mismatches] = arg
+  end
+  opts.on("--print-amplicon", "print out the sequence of the amplicon [default: #{options[:print_amplicon] }]") do
+    options[:print_amplicon] = true
+  end
+
+  # logger options
+  opts.separator "\nVerbosity:\n\n"
+  opts.on("-q", "--quiet", "Run quietly, set logging to ERROR level [default INFO]") {options[:log_level] = 'error'}
+  opts.on("--logger filename",String,"Log to file [default #{options[:logger]}]") { |name| options[:logger] = name}
+  opts.on("--trace options",String,"Set log level [default INFO]. e.g. '--trace debug' to set logging level to DEBUG"){|s| options[:log_level] = s}
+end; o.parse!
+if ARGV.length != 0 or options[:primer1].nil? or options[:primer2].nil? or options[:fasta].nil?
+  $stderr.puts o
+  exit 1
+end
+# Setup logging
+Bio::Log::CLI.logger(options[:logger]); Bio::Log::CLI.trace(options[:log_level]); log = Bio::Log::LoggerPlus.new(LOG_NAME); Bio::Log::CLI.configure(LOG_NAME)
+
+
+# make the primer set
+primer_set = Bio::Ipcress::PrimerSet.new options[:primer1], options[:primer2]
+
+to_gc_binary = lambda do |seq|
+  str = ''
+  seq.each_char do |char|
+    if %(G C).include?(char)
+      str="#{str}1"
+    else
+      str="#{str}0"
+    end
+  end
+  str
+end
+to_gc_count = lambda do |seq|
+  count = 0
+  seq.each_char do |char|
+    count += 1 if %(G C).include?(char)
+  end
+  count
+end
+
+# output characters of each hit
+headers = %w(
+target
+mismatches_fwd
+mismatches_rev
+length
+gc_of_forward_matching
+gc_of_reverse_matching
+gc_positions_of_forward_matching
+gc_positions_of_reverse_matching
+)
+headers += ['amplicon'] if options[:print_amplicon]
+puts headers.join("\t")
+
+options[:fasta].each do |fasta|
+  # run ipcress
+  mismatch_param = 3 #default to 3 so ipcress bug gets worked around, and filter later
+  mismatch_param = options[:num_mismatches] if options[:num_mismatches] > mismatch_param
+
+  results = Bio::Ipcress.run primer_set, fasta, :mismatches => mismatch_param
+
+  seqs = {}
+  if options[:print_amplicon]
+    Bio::FlatFile.foreach(fasta) do |e|
+      name = e.definition
+      seq = e.seq.seq
+      raise "Duplicate sequence name found: #{name}" if seqs.key?(name)
+      seqs[name] = seq
+    end
+  end
+
+  results.each do |res|
+    misses = res.recalculate_mismatches_from_alignments
+    next if misses.reduce(:+) > options[:num_mismatches]
+
+    to_print = [
+      res.target,
+      misses[0],
+      misses[1],
+      res.length,
+      to_gc_count.call(res.forward_matching_sequence),
+      to_gc_count.call(res.reverse_matching_sequence),
+      to_gc_binary.call(res.forward_matching_sequence),
+      to_gc_binary.call(res.reverse_matching_sequence),
+      ]
+    if options[:print_amplicon]
+      name = res.target.gsub(':filter(unmasked)','') #H509DRAFT_scaffold00021.21:filter(unmasked)
+      raise "Unable to find sequence name #{name} in fasta file, possible programming error" if !seqs.key?(name)
+      seq = seqs[name]
+      amplicon = seq[(res.start)...(res.start+res.length)]
+      if res.result_type == 'forward'
+        to_print += [amplicon]
+      elsif res.result_type == 'revcomp'
+        to_print += [Bio::Sequence::NA.new(amplicon).reverse_complement.to_s.upcase]
+      else
+        raise "Unexpected ipcress result type: #{res.result_type}"
+      end
+    end
+    puts to_print.join("\t")
+  end
+end

data/test/test-script.rb

@@ -0,0 +1,21 @@
+require 'helper'
+require 'bio-commandeer'
+
+class TestScript < Test::Unit::TestCase
+  should "test 1 result" do
+    path_to_script = File.join(File.dirname(__FILE__),'..','bin','pcr.rb')
+    path_to_data = File.join(File.dirname(__FILE__),'data')
+
+    expected = "target	mismatches_fwd	mismatches_rev	length	gc_of_forward_matching	gc_of_reverse_matching	gc_positions_of_forward_matching	gc_positions_of_reverse_matching
+gi|335929284|gb|JN048683.1|:filter(unmasked) Methanocella conradii HZ254 16S ribosomal RNA gene, partial sequence	0	3	418	6	7	11010101100	00110011111
+gi|335929284|gb|JN048683.1|:filter(unmasked) Methanocella conradii HZ254 16S ribosomal RNA gene, partial sequence	0	3	733	6	5	11010101100	00111100001
+gi|335929284|gb|JN048683.1|:filter(unmasked) Methanocella conradii HZ254 16S ribosomal RNA gene, partial sequence	3	0	348	7	10	11010101101	1000010110001001100111
+gi|335929284|gb|JN048683.1|:filter(unmasked) Methanocella conradii HZ254 16S ribosomal RNA gene, partial sequence	3	0	123	8	10	11011101101	1000010110001001100111
+gi|335929284|gb|JN048683.1|:filter(unmasked) Methanocella conradii HZ254 16S ribosomal RNA gene, partial sequence	0	3	418	6	7	11010101100	00110011111
+gi|335929284|gb|JN048683.1|:filter(unmasked) Methanocella conradii HZ254 16S ribosomal RNA gene, partial sequence	0	3	733	6	5	11010101100	00111100001
+gi|335929284|gb|JN048683.1|:filter(unmasked) Methanocella conradii HZ254 16S ribosomal RNA gene, partial sequence	3	0	348	7	10	11010101101	1000010110001001100111
+gi|335929284|gb|JN048683.1|:filter(unmasked) Methanocella conradii HZ254 16S ribosomal RNA gene, partial sequence	3	0	123	8	10	11011101101	1000010110001001100111
+"
+    assert_equal(expected, Bio::Commandeer.run("#{path_to_script} --primer1 GGTCACTGCTA --primer2 GGCTACCTTGTTACGACTTAAC --fasta #{path_to_data}/Ipcress/Methanocella_conradii_16s_twice.fa"))
+  end
+end

metadata

@@ -1,109 +1,123 @@
 --- !ruby/object:Gem::Specification
 name: bio-ipcress
 version: !ruby/object:Gem::Version
-  version: 0.0.1
-  prerelease: 
+  version: 0.1.0
 platform: ruby
 authors:
 - Ben J Woodcroft
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-08-08 00:00:00.000000000 Z
+date: 2017-01-09 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bio
-  requirement: &72407200 !ruby/object:Gem::Requirement
-    none: false
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: 1.4.2
   type: :runtime
   prerelease: false
-  version_requirements: *72407200
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.4.2
 - !ruby/object:Gem::Dependency
   name: shoulda
-  requirement: &72406750 !ruby/object:Gem::Requirement
-    none: false
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *72406750
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: rdoc
-  requirement: &72406370 !ruby/object:Gem::Requirement
-    none: false
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '3.12'
   type: :development
   prerelease: false
-  version_requirements: *72406370
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.12'
 - !ruby/object:Gem::Dependency
   name: jeweler
-  requirement: &72406020 !ruby/object:Gem::Requirement
-    none: false
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.3.2
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.8.3
+        version: '2.3'
   type: :development
   prerelease: false
-  version_requirements: *72406020
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.3.2
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.3'
 - !ruby/object:Gem::Dependency
   name: bundler
-  requirement: &72405780 !ruby/object:Gem::Requirement
-    none: false
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: 1.0.21
   type: :development
   prerelease: false
-  version_requirements: *72405780
-- !ruby/object:Gem::Dependency
-  name: bio
-  requirement: &72405540 !ruby/object:Gem::Requirement
-    none: false
+  version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
-        version: 1.4.2
-  type: :development
-  prerelease: false
-  version_requirements: *72405540
+        version: 1.0.21
 - !ruby/object:Gem::Dependency
-  name: rdoc
-  requirement: &72405300 !ruby/object:Gem::Requirement
-    none: false
+  name: test-unit
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '3.12'
+        version: '0'
   type: :development
   prerelease: false
-  version_requirements: *72405300
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description: a programmatic interface to the iPCRess in-silico PCR software. iPCRess
   is part of the exonerate suite.
 email: gmail.com after donttrustben
-executables: []
+executables:
+- pcr.rb
 extensions: []
 extra_rdoc_files:
 - LICENSE.txt
 - README.md
 files:
-- .document
-- .travis.yml
+- ".document"
+- ".travis.yml"
 - Gemfile
 - LICENSE.txt
 - README.md
 - Rakefile
 - VERSION
+- bin/pcr.rb
 - lib/bio-ipcress.rb
 - lib/bio/appl/ipcress.rb
 - test/data/Ipcress/Methanocella_conradii_16s.fa
@@ -113,33 +127,30 @@ files:
 - test/data/Ipcress/input1_with_insert.txt
 - test/data/Ipcress/input2.txt
 - test/helper.rb
+- test/test-script.rb
 - test/test_bio-ipcress.rb
 homepage: http://github.com/wwood/bioruby-ipcress
 licenses:
 - MIT
+metadata: {}
 post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
-      segments:
-      - 0
-      hash: 237777805
 required_rubygems_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.17
+rubygems_version: 2.5.2
 signing_key: 
-specification_version: 3
+specification_version: 4
 summary: a programmatic interface to the iPCRess in-silico PCR software
 test_files: []