assemblotron 0.0.2

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 714db5f952949fb16921c707434b7ba07797c3aa
+  data.tar.gz: 407ce9019b37b5492a41d4b72e0101ea57f52c0e
+SHA512:
+  metadata.gz: e586502ee078aaee565064d21adc443ffff78bed298ed4a40b3b01208077129e738d3d126115597324551fe6767a99f70760e12fc4352214705373d5ac8d9150
+  data.tar.gz: b6e575f209ca04870fe1589dc3cdba1ce061664f859130889c5893c826560435dd7038c09e9d06272d14ea100ff73d220605b0cdccb8bcc55022cc048ce6b717

data/README.md

@@ -0,0 +1,39 @@
+assemblotron
+============
+
+Automated optimal *de-novo* assembly.
+
+Transcriptome assembly takes a *long* time and a *lot* of computational resources. The software is complex, and the best settings to use depend heavily on the organism being studied.
+
+Assemblotron solves this problem by rapidly discovering the optimal settings for an assembler or assembly pipeline and performs the best possible assembly using the tools available.
+
+A typical Assemblotron run takes only 3-4 hours on 8 cores of a desktop PC with an i7 processor, and greatly improves the accuracy of expression quantification and gene reconstruction in de-novo transcriptome analysis (blog posts/paper with evidence to follow shortly).
+
+## Explanation
+
+Assemblotron takes a small subsample of the available reads and runs an assembly. The assembly is thoroughly analysed and scored using [transrate](https://github.com/Blahah/transrate). Then the optimisation system [biopsy](https://github.com/Blahah/biopsy) is used to select new assembler settings to test, and another assembly is performed. This process is repeated until an estimate for the best possible assembly is found.
+
+Further documentation will be provided when the software enters beta.
+
+## Development status
+
+[![Gem Version](https://badge.fury.io/rb/assemblotron.png)][gem]
+[![Build Status](https://secure.travis-ci.org/Blahah/assemblotron.png?branch=master)][travis]
+[![Dependency Status](https://gemnasium.com/Blahah/assemblotron.png?travis)][gemnasium]
+[![Code Climate](https://codeclimate.com/github/Blahah/assemblotron.png)][codeclimate]
+[![Coverage Status](https://coveralls.io/repos/Blahah/assemblotron/badge.png?branch=master)][coveralls]
+
+[gem]: https://badge.fury.io/rb/assemblotron
+[travis]: https://travis-ci.org/Blahah/assemblotron
+[gemnasium]: https://gemnasium.com/Blahah/assemblotron
+[codeclimate]: https://codeclimate.com/github/Blahah/assemblotron
+[coveralls]: https://coveralls.io/r/Blahah/assemblotron
+
+This software is in pre-alpha development and is not yet ready for deployment. 
+Please don't report issues or request documentation until we are ready for beta release (see below for estimated timeframe).
+
+## Expectations for imminent versions
+
+* **v0.1.0**: allow optimisation of SoapDenovoTrans, VelvetOases (due 10th November 2013)
+* **v0.2.0**: add Abyss, SGA and ReadJoiner
+* **v0.3.0**: allow optimising pipelines

data/Rakefile

@@ -0,0 +1,8 @@
+require 'rake/testtask'
+
+Rake::TestTask.new do |t|
+  t.libs << 'test'
+end
+
+desc "Run tests"
+task :default => :test

data/lib/assemblotron.rb

@@ -0,0 +1,210 @@
+require 'biopsy'
+require 'logger'
+require 'transrate'
+require 'assemblotron/version'
+require 'assemblotron/sample'
+require 'pp'
+require 'json'
+
+module Assemblotron
+
+  include Transrate
+
+
+  class Controller
+  
+    attr_accessor :global_opts
+    attr_accessor :assembler_opts
+
+    # Return a new Assemblotron
+    def initialize
+      @log = Logger.new(STDOUT)
+      @log.level = Logger::INFO
+      self.load_config
+      self.init_settings
+      @assemblers = []
+      self.load_assemblers
+    end # initialize
+
+    def self.header
+      "Assemblotron v#{VERSION::STRING.dup}"
+    end
+
+    # Initialise the Biopsy settings
+    def init_settings
+      s = Biopsy::Settings.instance
+      s.set_defaults
+      libdir = File.dirname(__FILE__)
+      s.target_dir = [File.join(libdir, 'assemblotron/assemblers/')]
+      s.objectives_dir = [File.join(libdir, 'assemblotron/objectives/')]
+      @log.debug "initialised Biopsy settings"
+    end # init_settings
+
+    # Load global configuration from the config file at
+    # +~/.assemblotron+, if it exists.
+    def load_config
+      config_file = File.join(Dir.home, ".assemblotron")
+      if File.exists? config_file
+        @log.debug "config file found at #{config_file}"
+        config = YAML::load_file(config_file)
+        if config.nil?
+          @log.warn 'config file malformed or empty'
+          return
+        end
+        @config = config.deep_symbolize
+      end
+    end # parse_config
+
+    # Discover and load available assemblers.
+    #
+    # Loads all assemblers provided by the program, and
+    # then searches any directories listed in the config
+    # file (+~/.assemblotron+) setting +assembler_dirs+.
+    #
+    # Directories listed in +assembler_dirs+ must contain:
+    #
+    # +definitions+::  Directory with one +.yml+ definition per assembler.
+    #                  See the documentation for Definition.
+    # +constructors+:: Directory with one +.rb+ file per assembler.
+    #                  See the documentation for Constructor. 
+    def load_assemblers
+      Biopsy::Settings.instance.target_dir.each do |dir|
+        Dir.chdir dir do
+          Dir['*.yml'].each do |file|
+            name = File.basename(file, '.yml')
+            target = Biopsy::Target.new
+            target.load_by_name name
+            @assemblers << target
+          end
+        end
+      end
+    end # load_assemblers
+
+    # Return an array of the names of available assemblers
+    def assemblers
+      a = []
+      @assemblers.each do |t|
+        a << t.name
+        a << t.shortname if t.shortname
+      end
+      a
+    end # assemblers
+
+    def list_assemblers
+      puts Controller.header
+      puts <<-EOS
+
+Available assemblers are listed below.
+Shortnames are shown in brackets if available.
+
+Usage:
+atron [global options] <assembler> [assembler options]
+
+Assemblers:
+EOS
+      @assemblers.each do |a| 
+        p = " - #{a.name}"
+        p += " (#{a.shortname})" if a.respond_to? :shortname
+        puts p
+      end
+    end # list_assemblers
+
+    def options_for_assembler assembler
+      a = self.get_assembler assembler
+      parser = Trollop::Parser.new do
+          banner <<-EOS
+#{Controller.header}
+
+Options for assembler #{assembler}
+EOS
+        opt :reference, "Path to reference proteome file in FASTA format",
+             :type => String,
+             :required => true
+        a.options.each_pair do |param, opts|
+          opt param, 
+              opts[:desc], 
+              :type => Controller.class_from_type(opts[:type])
+        end
+      end
+      Trollop::with_standard_exception_handling parser do
+        raise Trollop::HelpNeeded if ARGV.empty? # show help screen
+        parser.parse ARGV
+      end
+    end # options_for_assembler
+
+    def get_assembler assembler
+      ret = @assemblers.find do |a|
+        a.name == assembler || 
+        a.shortname == assembler
+      end
+      raise "couldn't find assembler #{assembler}" if ret.nil?
+      ret
+    end
+
+    def self.class_from_type type
+      case type
+      when 'string'
+        String
+      when 'int'
+        Integer
+      when 'float'
+        Float
+      end
+    end
+
+    def subsample_input
+      l = @assembler_opts[:left]
+      r = @assembler_opts[:right]
+      size = @global_opts[:subsample_size]
+
+      s = Sample.new(l, r)
+      ls, rs = s.subsample size
+ 
+      @assembler_opts[:left_subset] = ls
+      @assembler_opts[:right_subset] = rs
+    end
+
+    def final_assembly assembler, result
+      Dir.mkdir('final_assembly')
+      Dir.chdir('final_assembly') do
+        assembler.run result
+      end
+    end
+
+    def run assembler
+      # subsampling
+      if @global_opts[:skip_subsample]
+        @assembler_opts[:left_subset] = assembler_opts[:left]
+        @assembler_opts[:right_subset] = assembler_opts[:right]
+      else
+        subsample_input
+      end
+
+      # load reference and create ublast DB
+      @assembler_opts[:reference] = Transrate::Assembly.new(@assembler_opts[:reference])
+      ra = Transrate::ReciprocalAnnotation.new(@assembler_opts[:reference], @assembler_opts[:reference])
+      ra.make_reference_db
+
+      # setup the assembler
+      a = self.get_assembler assembler
+      a.setup_optim(@global_opts, @assembler_opts)
+
+      # run the optimisation
+      e = Biopsy::Experiment.new(a, options: @assembler_opts, threads: @global_opts[:threads])
+      res = e.run
+
+      # write out the result
+      File.open(@global_opts[:output_parameters], 'wb') do |f|
+        f.write(JSON.pretty_generate(res))
+      end
+
+      # run the final assembly
+      a.setup_final(@global_opts, @assembler_opts)
+      unless @global_opts[:skip_final]
+        final_assembly a, res
+      end
+    end # run
+
+  end # Controller
+
+end # Assemblotron

data/lib/assemblotron/assemblers/constructors/abyss.rb

@@ -0,0 +1,7 @@
+class Abyss
+
+  def run params
+
+  end
+
+end # Abyss

data/lib/assemblotron/assemblers/constructors/read_joiner.rb

@@ -0,0 +1,7 @@
+class ReadJoiner
+
+  def run params
+
+  end
+
+end # ReadJoiner

data/lib/assemblotron/assemblers/constructors/sga.rb

@@ -0,0 +1,7 @@
+class SGA
+
+  def run params
+
+  end
+
+end # SGA

data/lib/assemblotron/assemblers/constructors/soap_denovo_trans.rb

@@ -0,0 +1,69 @@
+class SoapDenovoTrans
+
+  def initialize
+    @count = 0
+  end
+
+  def run params
+    self.setup_soap(params) if @count == 0
+    self.run_soap params
+    @count += 1
+  end
+
+  # soapdt.config file only generated on first run
+  def setup_soap params
+    # make config file
+    rf = params[:readformat] == 'fastq' ? 'q' : 'f'
+    File.open("soapdt.config", "w") do |conf|
+      conf.puts "max_rd_len=20000"
+      conf.puts "[LIB]"
+      conf.puts "avg_ins=#{params[:insertsize]}"
+      conf.puts "reverse_seq=0"
+      conf.puts "asm_flags=3"
+      conf.puts "rank=2"
+      conf.puts "#{rf}1=#{params[:l]}"
+      conf.puts "#{rf}2=#{params[:r]}"
+    end
+  end
+
+  def include_defaults params
+    defaults = {
+      :K => 23,
+      :p => 8,
+      :d => 0,
+      :e => 2,
+      :M => 1,
+      :F => true,
+      :L => 100,
+      :t => 5,
+      :G => 50
+    }
+    defaults.merge params
+  end
+
+  def construct_command(params)
+    params = self.include_defaults params
+    cmd = "#{params[:path]} all"
+    # generic
+    cmd += " -s soapdt.config" # config file
+    cmd += " -a #{params[:memory]}" # memory assumption
+    cmd += " -o #{params[:out]}" # output directory
+    cmd += " -p #{params[:threads]}" # number of threads
+    # specific
+    cmd += " -K #{params[:K]}" # kmer size
+    cmd += " -d #{params[:d]}" # minimum kmer frequency
+    cmd += " -F" if params[:F] # fill gaps in scaffold
+    cmd += " -M #{params[:M]}" # strength of contig flattening
+    cmd += " -L #{params[:L]}" # minimum contig length
+    cmd += " -e #{params[:e]}" # delete contigs with coverage no greater than
+    cmd += " -t #{params[:t]}" # maximum number of transcripts from one locus
+    cmd += " -G #{params[:G]}" # allowed length difference between estimated and filled gap
+  end
+
+  # runs SOAPdt script
+  def run_soap(params)
+    cmd = self.construct_command(params)
+    `#{cmd} > #{@count}.log`
+  end
+
+end # SoapDenovoTrans

data/lib/assemblotron/assemblers/constructors/velvet_oases.rb

@@ -0,0 +1,7 @@
+class VelvetOases
+
+  def run params
+
+  end
+
+end # VelvetOases

data/lib/assemblotron/assemblers/soap_denovo_trans.rb

@@ -0,0 +1,120 @@
+require 'transrate'
+
+class SoapDenovoTrans
+
+  include Which
+
+  # Return a new SoapDenovoTrans object
+  def initialize
+    @count = 0
+    @path = which('SOAPdenovo-Trans-127mer')
+    raise "SOAPdenovo-Trans-127mer was not in the PATH" if @path.empty?
+    @path = @path.first
+  end
+
+  # Run the assembler with the provided parameters,
+  # returning a Transrate::ComparativeMetrics object
+  # containing a score for the generated assembly
+  # compared to the reference.
+  def run params
+    # run the assembly
+    self.run_soap params
+    # retrieve output
+    scaffolds = Dir['*.scafSeq']
+    return nil if scaffolds.empty?
+    scaffolds = scaffolds.first
+    return nil if File.size(scaffolds) == 0
+    # return a ComparativeMetrics object
+    assembly = Transrate::Assembly.new(scaffolds)
+    Transrate::ComparativeMetrics.new(assembly, params[:reference])
+  end
+
+  # Perform any necessary setup for the assembler
+  # prior to running the parameter optimisation.
+  # This includes modifying the config file to point
+  # to the subsetted reads rather than the full set.
+  def setup_optim(global_opts, assembler_opts)
+    # setup config file for subsetted reads
+    left = assembler_opts[:left_subset]
+    right = assembler_opts[:right_subset]
+    f = create_config(left, right, assembler_opts)
+    assembler_opts[:config] = f
+  end
+
+  # Perform any necessary setup for the assembler
+  # prior to running the full optimal assembly.
+  # This includes resetting the config to refer
+  # to the full set of reads.
+  def setup_full(global_opts, assembler_opts)
+    # set config file for full read set
+    left = assembler_opts[:left]
+    right = assembler_opts[:right]
+    f = create_config(left, right, assembler_opts)
+    assembler_opts[:config] = f
+  end
+
+  # Generate a config file with the specified left and right
+  # read input files, returning the full path to the config file.
+  def create_config left, right, assembler_opts
+    # create the config file
+    filename = "#{Time.now}.full.config"
+    File.open(filename) do |f|
+      f << 'max_rd_len=5000'
+      f << '[LIB]'
+      f << "avg_ins=#{assembler_opts[:insertsize]}"
+      f << "reverse_seq=0" # don't reverse complement the reads
+      f << "asm_flags=3"   # use the reads for assembly and scaffolding
+      f << "q1=#{left}"
+      f << "q2=#{right}"
+    end
+    File.expand_path filename
+  end
+
+  # Merge the default parameters with the hash provided
+  def include_defaults params
+    defaults = {
+      :K => 23,
+      :threads => 8,
+      :out => 'sdt',
+      :d => 0,
+      :e => 2,
+      :M => 1,
+      :F => true,
+      :L => 100,
+      :t => 5,
+      :G => 50
+    }
+    defaults.merge params
+  end
+
+  # Given a set of parameters, fill in any missing
+  # parameters with defaults and construct a command
+  # to run the target assembler. Return the command
+  # as a string.
+  def construct_command params
+    params = self.include_defaults params
+    cmd = "#{@path} all"
+    # generic
+    cmd += " -s #{params[:config]}" # config file
+    cmd += " -a #{params[:memory]}" if params.has_key? :memory # memory assumption
+    cmd += " -o #{params[:out]}" if params.has_key? :out # output directory
+    cmd += " -p #{params[:threads]}" # number of threads
+    # specific
+    cmd += " -K #{params[:K]}" # kmer size
+    cmd += " -d #{params[:d]}" # minimum kmer frequency
+    cmd += " -F" if params[:F] # fill gaps in scaffold
+    cmd += " -M #{params[:M]}" # strength of contig flattening
+    cmd += " -L #{params[:L]}" # minimum contig length
+    cmd += " -e #{params[:e]}" # delete contigs with coverage no greater than
+    cmd += " -t #{params[:t]}" # maximum number of transcripts from one locus
+    cmd += " -G #{params[:G]}" # allowed length difference between estimated and filled gap
+  end
+
+  # Run the SOAPdenovo-trans assembler with the provided
+  # parameters. Return the output generated.
+  def run_soap(params)
+    cmd = self.construct_command(params)
+    `#{cmd} > #{@count}.log`
+  end
+
+end # SoapDenovoTrans