#!/usr/bin/env ruby
-# Copyright (C) 2007-2010 Martin A. Hansen.
+# Copyright (C) 2007-2011 Martin A. Hansen.
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License
# >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> DESCRIPTION <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
-#
+# Run Uclust on sequences in the stream.
# >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
+require 'maasha/biopieces'
+require 'maasha/fasta'
+require 'maasha/usearch'
-require 'biopieces'
-require 'fasta'
-
-class Uclust
- include Enumerable
-
- def initialize(infile, outfile)
- @infile = infile
- @outfile = outfile
- end
-
- # Method that calls Uclusts sorting for sorting a FASTA file
- # according to decending sequence length.
- def sort(options)
- command = "nice -n 19"
- command << " uclust --sort #{@infile} --output #{@infile}.sort"
- command << " > /dev/null 2>&1" unless options[:verbose]
- system(command)
- raise "Failed sorting file: #{@infile}" unless $?.success?
-
- File.rename "#{@infile}.sort", @infile
- end
-
- def ublast(options)
- # uclust --ublast query_seqs.fasta --db database.fasta --blast6out filename --evalue E
- options[:e_val] = 10 if options[:e_val].is_nil?
- command = "nice -n 19"
- command << " uclust --ublast #{@infile} --db #{options[:database]} --uc #{@outfile} --evalue #{options[:e_val]}"
- command << " > /dev/null 2>&1" unless options[:verbose]
- system(command)
- raise "Command failed: #{command}" unless $?.success?
- end
+casts = []
+casts << {:long=>'no_sort', :short=>'n', :type=>'flag', :mandatory=>false, :default=>nil, :allowed=>nil, :disallowed=>nil}
+casts << {:long=>'comp', :short=>'c', :type=>'flag', :mandatory=>false, :default=>nil, :allowed=>nil, :disallowed=>nil}
+casts << {:long=>'identity', :short=>'i', :type=>'float', :mandatory=>true, :default=>0.9, :allowed=>nil, :disallowed=>nil}
- def usearch(options)
- # uclust --query query.fasta --db db.fasta --uc results.uc --id 0.90 [--evalue E]
- command = "nice -n 19"
- command << " uclust --query #{@infile} --db #{options[:database]} --uc #{@outfile} --id #{options[:identity]}"
- command << " --evalue #{options[:e_val]}" if options.has_key? :e_val
- command << " > /dev/null 2>&1" unless options[:verbose]
- system(command)
- raise "Command failed: #{command}" unless $?.success?
- end
+options = Biopieces.options_parse(ARGV, casts)
- def uclust(options)
- # uclust --input seqs_sorted.fasta --uc results.uc --id 0.90
- command = "nice -n 19"
- command << " uclust --input #{@infile} --uc #{@outfile} --id #{options[:identity]}"
- command << " > /dev/null 2>&1" unless options[:verbose]
- system(command)
- raise "Command failed: #{command}" unless $?.success?
- end
+tmpdir = Biopieces.mktmpdir
+file_records = File.join(tmpdir, "data.stream")
+file_fasta = File.join(tmpdir, "in.fna")
+file_uclust = File.join(tmpdir, "out.uc")
- def usearch_uclust(options)
- # uclust --input seqs_sorted.fasta --lib db.fasta --uc results.uc --id 0.90
- command = "nice -n 19"
- command << " uclust --input #{@infile} --lib #{options[:database]} --uc #{@outfile} --id #{options[:identity]}"
- command << " --lib #{options[:database]}" if options.has_key? :database
- command << " > /dev/null 2>&1" unless options[:verbose]
- system(command)
- raise "Command failed: #{command}" unless $?.success?
- end
+Biopieces.open(options[:stream_in], file_records) do |input, output|
+ Fasta.open(file_fasta, mode="w") do |fasta_io|
+ input.each_record do |record|
+ output.puts record
- def each
- record = {}
-
- File.open(@outfile, mode="r") do |ios|
- ios.each_line do |line|
- if line !~ /^#/
- fields = line.split("\t")
-
- record[:REC_TYPE] = "UCLUST"
- record[:TYPE] = fields[0]
- record[:CLUSTER] = fields[1]
- record[:SEQ_LEN] = fields[2]
- record[:IDENT] = fields[3]
- record[:STRAND] = fields[4]
- record[:Q_BEG] = fields[5]
- record[:S_BEG] = fields[6]
- record[:CIGAR] = fields[7]
- record[:Q_ID] = fields[8]
- record[:S_ID] = fields[9]
-
- yield record
- end
+ if record.has_key? :SEQ_NAME and record.has_key? :SEQ
+ fasta_io.puts Seq.new_bp(record).to_fasta
end
end
-
- self # conventionally
end
end
-ok_methods = "ublast,usearch,uclust,usearch_uclust"
-
-casts = []
-casts << {:long=>'no_sort', :short=>'n', :type=>'flag', :mandatory=>false, :default=>nil, :allowed=>nil, :disallowed=>nil}
-casts << {:long=>'method', :short=>'m', :type=>'string', :mandatory=>true, :default=>"uclust", :allowed=>ok_methods, :disallowed=>nil}
-casts << {:long=>'database', :short=>'d', :type=>'file!', :mandatory=>false, :default=>nil, :allowed=>nil, :disallowed=>nil}
-casts << {:long=>'identity', :short=>'i', :type=>'float', :mandatory=>true, :default=>0.9, :allowed=>nil, :disallowed=>nil}
-casts << {:long=>'e_val', :short=>'e', :type=>'float', :mandatory=>false, :default=>nil, :allowed=>nil, :disallowed=>nil}
-
-bp = Biopieces.new
-
-options = bp.parse(ARGV, casts)
+us = Usearch.new(file_fasta, file_uclust, options)
+us.sort_length unless options[:no_sort]
+us.cluster
-tmpdir = bp.mktmpdir
-infile = "#{tmpdir}/in.fna"
-outfile = "#{tmpdir}/out.uc"
+hash = {}
-Fasta.open(infile, mode="w") do |fasta_io|
- bp.each_record do |record|
- bp.puts record
- fasta_io.puts record
- end
+us.each_cluster do |record|
+ hash[record[:Q_ID].to_sym] = record.dup
end
-uclust = Uclust.new(infile, outfile)
-uclust.sort(options) unless options[:no_sort]
-
-case options[:method].to_s
-when "ublast" then uclust.ublast(options)
-when "usearch" then uclust.usearch(options)
-when "uclust" then uclust.uclust(options)
-when "usearch_uclust" then uclust.usearch_uclust(options)
-else raise "Unknown method: #{options[:method]}"
-end
+Biopieces.open(file_records, options[:stream_out]) do |input, output|
+ input.each_record do |record|
+ if record.has_key? :SEQ_NAME and record.has_key? :SEQ
+ if hash.has_key? record[:SEQ_NAME].to_sym
+ us = hash[record[:SEQ_NAME].to_sym]
+ record[:CLUSTER] = us[:CLUSTER].to_i
+ record[:IDENT] = us[:IDENT].to_i
+ record[:IDENT] = '*' if us[:TYPE] == 'S'
+ end
+ end
-uclust.each do |record|
- bp.puts record
+ output.puts record
+ end
end