bio-gadget 0.4.8 → 0.5.0

data/test/bio/gadget_test.rb

@@ -0,0 +1,11 @@
+require 'test_helper'
+
+class Bio::GadgetTest < Minitest::Test
+  def test_that_it_has_a_version_number
+    refute_nil ::Bio::Gadget::VERSION
+  end
+
+  def test_it_does_something_useful
+    assert false
+  end
+end

data/test/test_helper.rb

@@ -0,0 +1,4 @@
+$LOAD_PATH.unshift File.expand_path('../../lib', __FILE__)
+require 'bio/gadget'
+
+require 'minitest/autorun'

metadata

@@ -1,114 +1,181 @@
 --- !ruby/object:Gem::Specification
 name: bio-gadget
 version: !ruby/object:Gem::Version
-  version: 0.4.8
+  version: 0.5.0
 platform: ruby
 authors:
 - Shintaro Katayama
 autorequire: 
-bindir: bin
+bindir: exe
 cert_chain: []
-date: 2013-06-09 00:00:00.000000000 Z
+date: 2017-02-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: thor
+  name: bundler
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.12'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.12'
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '10.0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '10.0'
+- !ruby/object:Gem::Dependency
+  name: rake-compiler
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  type: :runtime
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: parallel
+  name: minitest
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.0'
+- !ruby/object:Gem::Dependency
+  name: bio
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: levenshtein-ffi
+  name: damerau-levenshtein
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: bio-faster
+  name: mkfifo
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: mkfifo
+  name: parallel
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: thor
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.19.3
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.19.3
 description: Gadgets for bioinformatics
 email:
 - shintaro.katayama@gmail.com
 executables:
 - bio-gadget
-extensions: []
+- fq1l
+- rbg
+- strt
+extensions:
+- ext/bio_gadget/extconf.rb
 extra_rdoc_files: []
 files:
-- .gitignore
+- ".gitignore"
+- ".travis.yml"
 - Gemfile
-- Gthorfile
 - LICENSE
 - README.org
 - Rakefile
-- bin/bio-gadget
+- bin/console
+- bin/setup
 - bio-gadget.gemspec
-- lib/bio-gadget.rb
-- lib/bio-gadget/dedup.rb
-- lib/bio-gadget/demlt.rb
-- lib/bio-gadget/femrg.rb
-- lib/bio-gadget/fqxz.rb
-- lib/bio-gadget/peak.rb
-- lib/bio-gadget/qvstat.rb
-- lib/bio-gadget/rgt2mtx.rb
-- lib/bio-gadget/version.rb
-- lib/bio-gadget/wig5p.rb
-- lib/bio-gadget/wigchr.rb
+- exe/bio-gadget
+- exe/fq1l
+- exe/rbg
+- exe/strt
+- ext/bio_gadget/bio_gadget.c
+- ext/bio_gadget/bio_gadget.h
+- ext/bio_gadget/extconf.rb
+- lib/bio/gadget.rb
+- lib/bio/gadget/fq1l.rb
+- lib/bio/gadget/strt.rb
+- lib/bio/gadget/strt/count.rb
+- lib/bio/gadget/strt/depth.rb
+- lib/bio/gadget/strt/prepare_transcriptome.rb
+- lib/bio/gadgets.rb
+- test/bio/gadget_test.rb
+- test/test_helper.rb
 homepage: https://github.com/shka/ruby-bio-gadget
-licenses: []
+licenses:
+- MIT
 metadata: {}
 post_install_message: 
 rdoc_options: []
@@ -116,18 +183,20 @@ require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.0.0
+rubygems_version: 2.6.8
 signing_key: 
 specification_version: 4
-summary: Gadgets for bioinformatics
-test_files: []
+summary: ''
+test_files:
+- test/bio/gadget_test.rb
+- test/test_helper.rb

data/Gthorfile

@@ -1,2 +0,0 @@
-$:.unshift File.expand_path('../lib', __FILE__)
-require 'bio-gadget'

data/bin/bio-gadget

@@ -1,5 +0,0 @@
-#!/usr/bin/env ruby
-
-require 'bio-gadget'
-
-Bio::Gadget.start

data/lib/bio-gadget.rb

@@ -1,44 +0,0 @@
-require 'bio-gadget/version'
-require 'bio-gadget/dedup'
-require 'bio-gadget/demlt'
-require 'bio-gadget/femrg'
-require 'bio-gadget/fqxz'
-require 'bio-gadget/peak'
-require 'bio-gadget/qvstat'
-require 'bio-gadget/rgt2mtx'
-require 'bio-gadget/wig5p'
-require 'bio-gadget/wigchr'
-
-require 'tempfile'
-
-module Bio
-  class Gadget < Thor
-
-    private
-
-    def myopen(file, &block)
-      # how to write?
-      f = (/\|/ !~ file && /\.gz$/ =~ file) ? "| gunzip -c #{file}" : file
-      unless block.nil?
-        o = open(f); block.call(o); o.close
-      else
-        open(f)
-      end
-    end
-
-    @@mytemppaths = Array.new
-
-    def mytemppath(basename, tmpdir = Dir::tmpdir)
-      fp = Tempfile.open(basename, tmpdir)
-      path = fp.path
-      @@mytemppaths.push(path)
-      fp.close!
-      path
-    end
-
-    END {
-      @@mytemppaths.each { |path| File.unlink(path) if File.exist?(path) }
-    }
-
-  end
-end

data/lib/bio-gadget/dedup.rb

@@ -1,33 +0,0 @@
-require 'bio-faster'
-require 'parallel'
-
-module Bio
-  class Gadget < Thor
-    namespace :bio
-
-    desc 'dedup', 'deduplicate fastq (via STDIN)'
-    def dedup
-
-      p1in, p1out = IO.pipe
-
-      fork {
-        p1in.close
-        $stdout.reopen(p1out)
-        open("| sort -k 1 -r -S #{sprintf('%2d', 100/(Parallel.processor_count+1))}% -T $TMPDIR | cut -f 2- | uniq -f 2", 'w') { |fp|
-          Bio::Faster.new(:stdin).each_record(:quality => :raw) do |seqid, seq, qvs|
-            fp.puts "#{seq}#{qvs}\t#{seqid}\t#{qvs}\t#{seq}"
-          end
-        }
-      }
-
-      p1out.close
-
-      p1in.each_line { |line|
-        seqid, qvs, seq = line.rstrip.split
-        puts "@#{seqid}\n#{seq}\n+\n#{qvs}"
-      }
-
-    end
-
-  end
-end

data/lib/bio-gadget/demlt.rb

@@ -1,149 +0,0 @@
-require 'bio-faster'
-require 'levenshtein'
-require 'mkfifo'
-require 'parallel'
-
-module Bio
-  class Gadget < Thor
-
-    namespace :bio
-
-    desc 'demlt BARCODE [FASTQ]', "Demultiplex fastq from STDIN by barcodes.\n\n"
-    option 'output-dir', :aliases => '-o', :type => :string, :default => '.'
-    option 'umi-length', :aliases => '-u', :type => :numeric, :default => 4, :desc => '0 is no umi, means no PCR-amplicon reduction.'
-    option 'cdna-length', :aliases => '-c', :type => :numeric, :default => 37, :desc => 'Trimming length before PCA-amplicon reduction. -1 is no trimming by length.'
-    option 'g-trimming', :aliases => '-g', :type => :boolean, :default => false, :desc => "Trimming of 5'-end poly-G. Length of the trimmed Gs attached after the read name."
-    option 'q-trimming', :aliases => '-q', :type => :string, :default => '~', :desc => "Quality threshold - nucleotides with lower quality will be trimmed, from the end of the sequence. '~' is no trimming by quality, because this is the maximum quality base character."
-    option 'min-length', :aliases => '-l', :type => :numeric, :default => 0, :desc => 'Length threshold - sequences shorter than this after trimming will be filtered out. 0 is no filtering.'
-    def demlt(bcfile, fastq=:stdin)
-
-      ofs = options['umi-length']
-      clen = options['cdna-length']
-      gtrim = options['g-trimming']
-      qtrim = options['q-trimming']
-      mlen = options['min-length']
-
-      wells = Array.new
-      bcs = Array.new
-      bclens = Array.new
-      open(bcfile).each do |line|
-        cols = line.rstrip.split
-        wells.push(cols[0])
-        bcs.push(cols[1])
-        bclens.push(cols[1].length)
-      end
-
-      bclens.uniq!
-      if bclens.size != 1
-        raise 'Inconsistent barcode sequence lengths'
-      end
-      bclen = bclens[0]
-
-      procs = Parallel.processor_count
-
-      fifo1paths = Array.new
-      procs.times { |i|
-        fifo1path = mytemppath('fifo1-')
-        File.mkfifo(fifo1path)
-        fifo1paths.push(fifo1path)
-      }
-      pid = Kernel.fork {
-        fifo1s = Array.new
-        fifo1paths.each { |fifo1path| fifo1s.push(open(fifo1path, 'w')) }
-        total = 0
-        Bio::Faster.new(fastq).each_record(:quality => :raw) do |vals|
-          fifo1 = fifo1s[total % procs]
-          fifo1.puts(vals.join("\t"))
-          total += 1
-        end
-        fifo1s.each { |fifo1| fifo1.close }
-        Kernel.exit!
-      }
-
-      fifo2paths = Array.new
-      procs.times { |i|
-        fifo2path = mytemppath('fifo2-')
-        File.mkfifo(fifo2path)
-        fifo2paths.push(fifo2path)
-        pid = Kernel.fork {
-          open(fifo2path, 'w') { |fifo2|
-            open(fifo1paths[i], 'r').each { |line|
-              seqid, seq, qvs = line.rstrip.split(/\t/)
-              tmpdists = Hash.new
-              bcs.each_index { |bcidx|
-                tmpdists[bcidx] = Levenshtein.distance(bcs[bcidx], seq[ofs, bclen])
-              }
-              dists = tmpdists.sort { |a, b| a[1] <=> b[1] }
-              bc = dists[0][1] < 2 && dists[0][1] < dists[1][1] ? dists[0][0] : -1
-              fifo2.puts("#{bc}\t#{seqid}\t#{seq}\t#{qvs}")
-            }
-          }
-          Kernel.exit!
-        }
-      }
-
-      tmpwells = wells + ['other']
-
-      fifo3paths = Array.new
-      tmpwells.each_index { |i|
-        fifo3path = mytemppath('fifo3-')
-        File.mkfifo(fifo3path)
-        fifo3paths.push(fifo3path)
-      }
-      pid = Kernel.fork {
-        fifo2s = Array.new
-        fifo2paths.each { |fifo2path| fifo2s.push(open(fifo2path, 'r')) }
-        fifo2done = Hash.new
-        fifo3s = Array.new
-        fifo3paths.each { |fifo3path| fifo3s.push(open(fifo3path, 'w')) }
-        fifo2s.cycle { |fifo2|
-          unless fifo2done.key?(fifo2)
-            line = fifo2.gets
-            if line.nil?
-              fifo2done[fifo2] = ''
-            else
-              bcs, seqid, seq, qvs = line.rstrip.split(/\t/)
-              fifo3 = fifo3s[bcs.to_i]
-              fifo3.puts([seqid, seq, qvs].join("\t"))
-            end
-          end
-          if fifo2done.size == fifo2s.size
-            break
-          end
-        }
-        fifo2s.each { |fifo2| fifo2.close }
-        fifo3s.each { |fifo3| fifo3.close }
-        Kernel.exit!
-      }
-
-      tmpwells.each_index { |i|
-        well = tmpwells[i]
-        outpath = "#{options['output-dir']}/#{well}.fq.xz"
-        pid = Kernel.fork {
-          left = ofs+bclen
-          right = clen > -1 ? -1 : ofs+bclen+clen-1
-          preprocess = ofs > 0 ? <<"DEDUPandFORMAT"
-ruby -F'\\t' -anle 'f1=$F[1][0..#{right}];f2=$F[2][0..#{right}];puts([f1+f2, $F[0], f2, f1].join("\\t"))' #{fifo3paths[i]} \\
-| sort -k 1 -r | cut -f 2- | uniq -f 2 \\
-| ruby -F'\\t' -anle 'puts(["@"+$F[0], $F[2][#{left}..-1], "+", $F[1][#{left}..-1]].join("\\n"))' \\
-DEDUPandFORMAT
-          : <<"FORMAT"
-ruby -F'\\t' -anle 'puts(["@"+$F[0], $F[1][#{left}..#{right}], "+", $F[2][#{left}..#{right}].rstrip].join("\\n"))' #{fifo3paths[i]} \\
-FORMAT
-
-          preprocess += '| ruby -e \'require "bio-faster";Bio::Faster.new(:stdin).each_record(:quality=>:raw){|v|s=v[1].gsub(/^G+/,"");l=v[1].length-s.length;puts("@#{v[0]}|-G#{l}\\n#{s}\\n+\\n#{v[2][l,s.length]}") if s.length>0}\'' if gtrim
-
-          if qtrim != '~' || mlen > 0
-            preprocess += '| ruby -e \'require "bio-faster";Bio::Faster.new(:stdin).each_record(:quality=>:raw){|v|m=v[2].length-1;0.upto(m){|i|if v[2][i]<"'+qtrim+'" then m=i-1;break;end};puts("@#{v[0]}\n#{v[1][0..m]}\n+\n#{v[2][0..m]}") if m+1>='+mlen.to_s+'}\''
-          end
-
-          exec preprocess+"| xz -z -c -e > #{outpath}"
-        }
-      }
-
-      Process.waitall
-
-    end
-
-  end
-end