big_simon 0.0.1

data/Gemfile.lock

@@ -0,0 +1,41 @@
+PATH
+  remote: .
+  specs:
+    big_simon (0.0.1)
+      rya (~> 0.1.3)
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    abort_if (0.2.0)
+    diff-lcs (1.3)
+    rake (10.5.0)
+    rspec (3.7.0)
+      rspec-core (~> 3.7.0)
+      rspec-expectations (~> 3.7.0)
+      rspec-mocks (~> 3.7.0)
+    rspec-core (3.7.1)
+      rspec-support (~> 3.7.0)
+    rspec-expectations (3.7.0)
+      diff-lcs (>= 1.2.0, < 2.0)
+      rspec-support (~> 3.7.0)
+    rspec-mocks (3.7.0)
+      diff-lcs (>= 1.2.0, < 2.0)
+      rspec-support (~> 3.7.0)
+    rspec-support (3.7.1)
+    rya (0.1.3)
+      abort_if (~> 0.2.0)
+      systemu (~> 2.6, >= 2.6.5)
+    systemu (2.6.5)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  big_simon!
+  bundler (~> 1.16)
+  rake (~> 10.0)
+  rspec (~> 3.0)
+
+BUNDLED WITH
+   1.16.3

data/README.md

@@ -0,0 +1,37 @@
+# BigSimon
+
+A pipeline for finding hosts of viruses!
+
+## Installation
+
+Add this line to your application's Gemfile:
+
+```ruby
+gem 'big_simon'
+```
+
+And then execute:
+
+    $ bundle
+
+Or install it yourself as:
+
+    $ gem install big_simon
+
+## Usage
+
+TODO
+
+## Development
+
+After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake spec` to run the tests. You can also run `bin/console` for an interactive prompt that will allow you to experiment.
+
+To install this gem onto your local machine, run `bundle exec rake install`. To release a new version, update the version number in `version.rb`, and then run `bundle exec rake release`, which will create a git tag for the version, push git commits and tags, and push the `.gem` file to [rubygems.org](https://rubygems.org).
+
+## Contributing
+
+Bug reports and pull requests are welcome on GitHub at https://github.com/mooreryan/big_simon. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the [Contributor Covenant](http://contributor-covenant.org) code of conduct.
+
+## Code of Conduct
+
+Everyone interacting in the BigSimon project’s codebases, issue trackers, chat rooms and mailing lists is expected to follow the [code of conduct](https://github.com/[USERNAME]/big_simon/blob/master/CODE_OF_CONDUCT.md).

data/Rakefile

@@ -0,0 +1,6 @@
+require "bundler/gem_tasks"
+require "rspec/core/rake_task"
+
+RSpec::Core::RakeTask.new(:spec)
+
+task :default => :spec

data/big_simon.gemspec

@@ -0,0 +1,30 @@
+
+lib = File.expand_path("../lib", __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+require "big_simon/version"
+
+Gem::Specification.new do |spec|
+  spec.name          = "big_simon"
+  spec.version       = BigSimon::VERSION
+  spec.authors       = ["Ryan Moore"]
+  spec.email         = ["moorer@udel.edu"]
+
+  spec.summary       = %q{Viral host discovery pipeline.}
+  spec.description   = %q{Viral host discovery pipeline.}
+  spec.homepage      = "https://github.com/mooreryan/big_simon"
+
+  # Specify which files should be added to the gem when it is released.
+  # The `git ls-files -z` loads the files in the RubyGem that have been added into git.
+  spec.files         = Dir.chdir(File.expand_path('..', __FILE__)) do
+    `git ls-files -z`.split("\x0").reject { |f| f.match(%r{^(test|spec|features)/}) }
+  end
+  spec.bindir        = "exe"
+  spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
+  spec.require_paths = ["lib"]
+
+  spec.add_development_dependency "bundler", "~> 1.16"
+  spec.add_development_dependency "rake", "~> 10.0"
+  spec.add_development_dependency "rspec", "~> 3.0"
+
+  spec.add_runtime_dependency "rya", "~> 0.1.3"
+end

data/bin/console

@@ -0,0 +1,14 @@
+#!/usr/bin/env ruby
+
+require "bundler/setup"
+require "big_simon"
+
+# You can add fixtures and/or initialization code here to make experimenting
+# with your gem easier. You can also use a different console, if you like.
+
+# (If you use this, don't forget to add pry to your Gemfile!)
+# require "pry"
+# Pry.start
+
+require "irb"
+IRB.start(__FILE__)

data/bin/setup

@@ -0,0 +1,8 @@
+#!/usr/bin/env bash
+set -euo pipefail
+IFS=$'\n\t'
+set -vx
+
+bundle install
+
+# Do any other automated setup that you need to do here

data/lib/big_simon/version.rb

@@ -0,0 +1,3 @@
+module BigSimon
+  VERSION = "0.0.1"
+end

data/lib/big_simon.rb

@@ -0,0 +1,100 @@
+require "rya"
+
+require "big_simon/version"
+
+Time.extend Rya::CoreExtensions::Time
+Process.extend Rya::CoreExtensions::Process
+
+module BigSimon
+
+  # Project directories
+  ROOT       = File.join __dir__, ".."
+  BIN        = File.join ROOT, "vendor", "bin", "mac"
+  SPEC       = File.join ROOT, "spec"
+  TEST_FILES = File.join SPEC, "test_files"
+
+  class Parsers
+
+    def self.vir_host_matcher fname
+      hosts = nil
+
+      host_info = {}
+      File.open(fname, "rt").each_line.with_index do |line, idx|
+        line.chomp!
+        line.sub! /,$/, "" # git rid of trailing commas
+
+        if idx.zero?
+          stat, *hosts = line.split ","
+        else
+          ary   = line.split ","
+          virus = ary.shift
+
+          dists = ary.map.
+              with_index { |dist, idx| [hosts[idx], dist.to_f] }.
+              sort_by { |_, dist| dist }
+
+          best_host = dists[0][0]
+
+          host_info[virus] = {
+              best: best_host,
+              all:  dists
+          }
+        end
+      end
+
+      host_info
+    end
+  end
+
+  class Runners
+
+    # Runs the WIsH program
+    #
+    # @raise [AbortIf::Exit] if commands fail
+    def self.wish exe, vir_dir, host_dir, outdir, threads
+      model_dir = File.join outdir, "model"
+
+      FileUtils.mkdir_p model_dir
+
+      build_model = "#{exe} " \
+      "-t #{threads} " \
+      "-c build " \
+      "-g #{host_dir} " \
+      "-m #{model_dir}"
+
+      predict = "#{exe} " \
+      "-t #{threads} " \
+      "-c predict " \
+      "-g #{vir_dir} " \
+      "-m #{model_dir} " \
+      "-r #{outdir} -b"
+
+      Process.run_and_time_it! "Building model", build_model
+      Process.run_and_time_it! "Predicting host", predict
+
+      FileUtils.rm_r model_dir if Dir.exist? model_dir
+    end
+
+    def self.vir_host_matcher exe, vir_dir, host_dir, outdir
+      FileUtils.mkdir_p outdir
+
+      cmd = "python #{exe} " \
+      "-v #{vir_dir} " \
+      "-b #{host_dir} " \
+      "-o #{outdir} " \
+      "-d 1" # only compute d2star dissimilarity
+
+      Process.run_and_time_it! "Computing d2star dissimilarity", cmd
+
+      tmp_dir = File.join outdir, "tmp"
+      FileUtils.rm_r tmp_dir if Dir.exist? tmp_dir
+
+      bad_files = %w[d2star_k6_main.html hostTaxa.txt_new.txt]
+      bad_files.each do |fname|
+        path = File.join outdir, fname
+
+        FileUtils.rm path if File.exist? path
+      end
+    end
+  end
+end

data/vendor/bin/mac/vhm.py

@@ -0,0 +1,381 @@
+#!/usr/bin/env python
+import os,sys
+import optparse
+import subprocess
+from subprocess import call
+import time
+import platform
+import numpy
+
+
+prog_base = os.path.split(sys.argv[0])[1]
+
+parser = optparse.OptionParser()
+parser.add_option("-v", "--virusFaDir", action = "store", type = "string", dest = "virusFaDir",
+									help = "the directory to the folder containing bacteria virus fasta files")
+parser.add_option("-b", "--hostFaDir", action = "store", type = "string", dest = "hostFaDir",
+									help = "the directory to the folder containing bacteria host fasta files")
+parser.add_option("-o", "--out", action = "store", type = "string", dest = "outDir",
+									default='./', help = "output directory")
+parser.add_option("-t", "--taxa", action = "store", type = "string", dest = "hostTaxaFile",
+									help = "the host taxa file (including the path) ")
+parser.add_option("-d", "--d2star", action = "store", type = "string", dest = "onlyD2star",
+									default=0, help = "compute only d2star dissimilarity? 1 for yes, 0 for no")
+#parser.add_option("-u", "--continue", action = "store", type = "string", dest = "onlyComputeMeasure",
+#									default=0, help = "kmer count is ready only compute measures? 1 for yes, 0 for no")
+#parser.add_option("-k", "--kLen", action = "store", type = "string", dest = "kLen",
+#									help = "the length of k-tuple")
+
+(options, args) = parser.parse_args()
+if (options.virusFaDir is None or
+		options.hostFaDir is None) :
+	sys.stderr.write(prog_base + ": ERROR: missing required command-line argument")
+	filelog.write(prog_base + ": ERROR: missing required command-line argument")
+	parser.print_help()
+	sys.exit(0)
+
+
+## tmp file directory
+if not os.path.exists(options.outDir) :
+	os.makedirs(options.outDir)
+tmpDir = os.path.join(options.outDir, "tmp")
+if not os.path.exists(tmpDir) :
+	os.makedirs(tmpDir)
+
+## log file ##
+filelog = open(os.path.join(tmpDir, 'vhm.log'), 'w')
+
+## name length ##
+nameLen = 93 - len(options.outDir)
+#### possibly because of the kmercount folder name for each contig is too long?
+
+
+#################### 0: preparation ############################
+
+## path to the programs
+vhmPath = os.path.dirname(sys.argv[0])
+if len(vhmPath) == 0 :
+	vhmPath="./"
+#print vhmPath
+
+## kmer length and MC order
+kmax = 6
+order = 2
+
+## compile c++ code if not
+optSys = platform.system()
+if optSys == 'Linux' :
+	exePath = os.path.join(vhmPath, "bin", "linux64")
+elif optSys == 'Darwin' :
+	exePath = os.path.join(vhmPath, "bin", "macDarwin")
+elif optSys == 'Windows' :
+	exePath = os.path.join(vhmPath, "bin", "windows64")
+else :
+	#sys.stderr.write("WARNING: can't recognize the operating system" + optSys + " \n")
+	exePath = "unknown"
+
+
+
+## countKmer c++ code
+countKmerCpp = os.path.join(vhmPath, "countKmer.cpp")
+countKmerOut = os.path.join(vhmPath, "countKmer.out")
+if not os.path.exists(countKmerOut) :
+	## recognize OS, copy exe files to vhmPath
+	sys.stderr.write("WARNING: can't find the file " + countKmerOut + ", try to copy one from bin \n")
+	filelog.write("WARNING: can't find the file " + countKmerOut + ", try to copy one from bin \n")
+	preCountKmerOut = os.path.join(exePath, "countKmer.out")
+	if exePath != "unknown" :
+		if os.path.exists(preCountKmerOut) :
+			os.system("cp " + preCountKmerOut + " " + vhmPath)
+		else :
+			sys.stderr.write( "ERROR: can't find file " + preCountKmerOut + " \n Please run make under the main directory! \n" )
+			filelog.write( "ERROR: can't find file " + preCountKmerOut + " \n Please run make under the main directory! \n" )
+			sys.exit(0)
+	else :
+	## can't recognize OS, try to compile
+		sys.stderr.write("WARNING: can't recognize the operating system" + optSys + " \n")
+		sys.stderr.write( "Trying to compile..." + "\n")
+		filelog.write("WARNING: can't recognize the operating system" + optSys + " \n")
+		filelog.write( "Trying to compile..." + "\n")
+		if os.path.exists(countKmerCpp) :
+			os.system("g++ " + countKmerCpp + " -o " + countKmerOut + "  -std=c++0x" )
+		else :
+			sys.stderr.write( "ERROR: can't find file " + countKmerCpp + " \n Please run make under the main directory! \n" )
+			filelog.write( "ERROR: can't find file " + countKmerCpp + " \n Please run make under the main directory! \n" )
+			sys.exit(0)
+
+os.system("chmod 777 " + countKmerOut)
+filelog.flush()
+
+## computeMeasure c++ code
+computeMeasureCpp = os.path.join(vhmPath, "computeMeasure.cpp")
+computeMeasureOut = os.path.join(vhmPath, "computeMeasure.out")
+if not os.path.exists(computeMeasureOut) :
+	## recognize OS, copy exe files to vhmPath
+	sys.stderr.write("WARNING: can't find the file " + computeMeasureOut + ", try to copy one from bin \n")
+	filelog.write("WARNING: can't find the file " + computeMeasureOut + ", try to copy one from bin \n")
+	preComputeMeasureOut = os.path.join(exePath, "computeMeasure.out")
+	if exePath != "unknown" :
+		if os.path.exists(preComputeMeasureOut) :
+			os.system("cp " + preComputeMeasureOut + " " + vhmPath)
+		else :
+			sys.stderr.write( "ERROR: can't find file " + preComputeMeasureOut + " \n Please run make under the main directory! \n" )
+			filelog.write( "ERROR: can't find file " + preComputeMeasureOut + " \n Please run make under the main directory! \n" )
+			sys.exit(0)
+	else :
+		## can't recognize OS, try to compile
+		sys.stderr.write("WARNING: can't recognize the operating system" + optSys + " \n")
+		sys.stderr.write( "Trying to compile..." + "\n")
+		filelog.write("WARNING: can't recognize the operating system" + optSys + " \n")
+		filelog.write( "Trying to compile..." + "\n")
+		if os.path.exists(computeMeasureCpp) :
+			os.system("g++ " + computeMeasureCpp + " -o " + computeMeasureOut + "  -std=c++0x" )
+		else :
+			sys.stderr.write( "ERROR: can't find file " + computeMeasureCpp + " \n Please run make under the main directory! \n" )
+			filelog.write( "ERROR: can't find file " + computeMeasureCpp + " \n Please run make under the main directory! \n" )
+			sys.exit(0)
+
+os.system("chmod 777 " + computeMeasureOut)
+filelog.flush()
+
+## computeMeasure c++ code
+computed2starCpp = os.path.join(vhmPath, "computeMeasure_onlyd2star.cpp")
+computed2starOut = os.path.join(vhmPath, "computeMeasure_onlyd2star.out")
+if not os.path.exists(computed2starOut) :
+	## recognize OS, copy exe files to vhmPath
+	sys.stderr.write("WARNING: can't find the file " + computed2starOut + ", try to copy one from bin \n")
+	preComputed2starOut = os.path.join(exePath, "computeMeasure_onlyd2star.out")
+	if exePath != "unknown" :
+		if os.path.exists(preComputed2starOut) :
+			os.system("cp " + preComputed2starOut + " " + vhmPath)
+		else :
+			sys.stderr.write( "ERROR: can't find file " + preComputed2starOut + " \n Please run make under the main directory! \n" )
+			filelog.write( "ERROR: can't find file " + preComputed2starOut + " \n Please run make under the main directory! \n" )
+			sys.exit(0)
+	else :
+		## can't recognize OS, try to compile
+		sys.stderr.write("WARNING: can't recognize the operating system" + optSys + " \n")
+		sys.stderr.write( "Trying to compile..." + "\n")
+		filelog.write("WARNING: can't recognize the operating system" + optSys + " \n")
+		filelog.write( "Trying to compile..." + "\n")
+		if os.path.exists(computed2starCpp) :
+			os.system("g++ " + computed2starCpp + " -o " + computed2starOut + "  -std=c++0x" )
+		else :
+			sys.stderr.write( "ERROR: can't find file " + computed2starCpp + " \n Please run make under the main directory! \n" )
+			filelog.write( "ERROR: can't find file " + computed2starCpp + " \n Please run make under the main directory! \n" )
+			sys.exit(0)
+
+os.system("chmod 777 " + computed2starOut)
+filelog.flush()
+
+## if only compute d2star
+if int(options.onlyD2star) == 1 :
+	computeMeasureOut=computed2starOut
+else :
+	computeMeasureOut=computeMeasureOut
+
+
+## kmer count directory
+kmerCountPath = os.path.join(tmpDir, "KC")
+if not os.path.exists(kmerCountPath) :
+	os.makedirs(kmerCountPath)
+
+## virusFaList, hostFaList
+virusFaList = os.listdir(options.virusFaDir)
+#virusFaList = os.path.join(options.virusFaDir, os.listdir(options.virusFaDir))
+hostFaList = os.listdir(options.hostFaDir)
+
+## virus list file, host list file
+virusListFile = os.path.join(tmpDir, "virusList")
+hostListFile = os.path.join(tmpDir, "hostList")
+
+virusListFileWrite = open(virusListFile, 'w') ## make file blank
+virusListFileWrite.close()
+virusListFileWrite = open(virusListFile, 'a')
+
+hostListFileWrite = open(hostListFile, 'w') ## make file blank
+hostListFileWrite.close()
+hostListFileWrite = open(hostListFile, 'a')
+
+#time.sleep(6) # delays for 10 seconds
+
+################ 00: hostTaxa issues: ##########################
+############## 1. hostName=hostFileName(with extension) ######
+############## 2. hostTaxa should be no missing   ########
+
+#################### 00: if hostTaxa missing ###################
+if options.hostTaxaFile is None :
+	hostTaxaFile = os.path.join(options.outDir, "hostTaxa.txt_new.txt")
+	sys.stdout.write("WARNING: no hostTaxa file provided, creating a dummy one \n")
+	filelog.write("WARNING: no hostTaxa file provided, creating a dummy one \n")
+	hostTaxaFileWrite = open(hostTaxaFile, 'w') ## make file blank
+	hostTaxaFileWrite.close()
+	hostTaxaFileWrite = open(hostTaxaFile, 'a')
+	
+	hostTaxaFileWrite.write("hostNCBIName	hostName	hostSuperkingdom	hostPhylum	hostClass	hostOrder	hostFamily	hostGenus	hostSpecies\n")
+	for currentFileName in hostFaList :
+		if currentFileName.startswith('.') :
+			continue
+		if len(currentFileName) > nameLen :
+			sys.stdout.write( "WARNING: the file name has more than " + str(nameLen) + " letters! Use the first " + str(nameLen) + " letters as the name \n")
+			filelog.write( "WARNING: the file name has more than " + str(nameLen) + " letters! Use the first " + str(nameLen) + " letters as the name \n")
+			currentFileNameS = currentFileName[:nameLen]
+		else :
+			currentFileNameS = currentFileName
+		hostTaxaFileWrite.write(currentFileNameS)
+		for i in range(1,9) :
+			hostTaxaFileWrite.write("\t" + "unknown")
+		hostTaxaFileWrite.write("\n")
+	hostTaxaFileWrite.close()
+	options.hostTaxaFile = hostTaxaFile
+
+################### REFORMAT hostTaxa (fill missing) ###############
+else :
+	hostTaxaFile = os.path.join(options.outDir, os.path.basename(options.hostTaxaFile)+"_new.txt")
+	hostTaxaTable = numpy.genfromtxt(options.hostTaxaFile,delimiter="\t", dtype=str)
+	hostTaxaTable[hostTaxaTable=='']='unknown'
+	numpy.savetxt(hostTaxaFile, hostTaxaTable, fmt="%s", delimiter='\t', newline='\n')
+    
+filelog.flush()
+
+#################### 1: count kmer and prepare list files ############################
+#sys.stdout.write("Step 1: counting kmers \n")
+
+start_time = time.time()
+count = 0
+
+for currentFileName in virusFaList :
+	if currentFileName.startswith('.') :
+		continue
+	if os.path.isdir(os.path.join(options.virusFaDir, currentFileName)) :
+		sys.stderr.write( "ERROR: zero bytes of file " + currentFileName + "\n")
+		filelog.write( "ERROR: zero bytes of file " + currentFileName + "\n")
+		sys.exit(0)
+	if len(currentFileName) > nameLen :
+		currentFileNameS = currentFileName[:nameLen]
+	else :
+		currentFileNameS = currentFileName
+	sys.stdout.write("Step 1: counting kmers for virus " + currentFileNameS + "\n")
+	filelog.write("Step 1: counting kmers for virus " + currentFileNameS + "\n")
+	for w in range(1, (kmax+1)) :
+		currentFilePath = os.path.join(options.virusFaDir, currentFileName)
+    
+		currentKmerCountPath = os.path.join(kmerCountPath, currentFileNameS)
+		cmdKmer = countKmerOut + " -l -k " + str(w) + \
+										" -i " + currentFilePath +\
+										" -o " + currentKmerCountPath +\
+										" -s " + currentFileNameS
+		cmdKmerOut = subprocess.Popen(cmdKmer, shell=True, \
+																			stderr = subprocess.PIPE, \
+																			stdout = subprocess.PIPE)
+		cmdKmerOut.wait()
+
+	if len(os.listdir(currentKmerCountPath)) == ( kmax + 1 ):
+		virusListFileWrite.write(currentFileNameS + " " + \
+													 currentKmerCountPath + " " +\
+													str(2) + "\n")
+	else :
+		sys.stderr.write( "ERROR in counting kmers for " + currentFileNameS + "\n")
+		filelog.write( "ERROR in counting kmers for " + currentFileNameS + "\n")
+		sys.exit(0)
+  
+	filelog.flush()
+  
+	end_time = time.time()
+	count += 1
+	#sys.stdout.write(str(end_time - start_time) + "s for " + str(count) + " seqs \n")
+	sys.stdout.write(" (Average time for counting kmers for one virus: " + str(round((end_time - start_time)/count,4)) + "s) \n" )
+	sys.stdout.write(" (ETR for counting kmers for viruses: " + str(round((end_time - start_time)/count*(len(virusFaList)-count), 4)) + "s) \n")
+	filelog.write(" (Average time for counting kmers for one virus: " + str(round((end_time - start_time)/count,4)) + "s) \n" )
+	filelog.write(" (ETR for counting kmers for viruses: " + str(round((end_time - start_time)/count*(len(virusFaList)-count), 4)) + "s) \n")
+
+virusListFileWrite.close()
+
+
+start_time = time.time()
+count = 0
+
+for currentFileName in hostFaList :
+	if currentFileName.startswith('.') :
+		continue
+	if os.path.isdir(os.path.join(options.hostFaDir, currentFileName)) :
+		sys.stderr.write( "ERROR: zero bytes of file " + currentFileName + "\n")
+		filelog.write( "ERROR: zero bytes of file " + currentFileName + "\n")
+		sys.exit(0)
+	if len(currentFileName) > nameLen :
+		currentFileNameS = currentFileName[:nameLen]
+	else :
+		currentFileNameS = currentFileName
+	sys.stdout.write("Step 1: counting kmers for host " + currentFileNameS + "\n")
+	filelog.write("Step 1: counting kmers for host " + currentFileNameS + "\n")
+	for w in range(1, (kmax+1)) :
+		currentFilePath = os.path.join(options.hostFaDir, currentFileName)
+		currentKmerCountPath = os.path.join(kmerCountPath, currentFileNameS)
+		cmdKmer = countKmerOut + " -l -k " + str(w) + \
+							" -i " + currentFilePath +\
+							" -o " + currentKmerCountPath +\
+							" -s " + currentFileNameS
+		#print cmdKmer
+		cmdKmerOut = subprocess.Popen(cmdKmer, shell=True, \
+																			stderr = subprocess.PIPE, \
+																			stdout = subprocess.PIPE)
+		cmdKmerOut.wait()
+
+	if len(os.listdir(currentKmerCountPath)) == ( kmax + 1 ) :
+		hostListFileWrite.write(currentFileNameS + " " + \
+												currentKmerCountPath + " " +\
+												str(2) + "\n")
+	else :
+		sys.stderr.write( "ERROR in counting kmers for " + currentFileNameS + "\n")
+		filelog.write( "ERROR in counting kmers for " + currentFileNameS + "\n")
+		sys.exit(0)
+  
+	filelog.flush()
+  
+	end_time = time.time()
+	count += 1
+	#sys.stdout.write(str(end_time - start_time) + "s for " + str(count) + " seqs \n")
+	sys.stdout.write(" (Average time for counting kmers for one host: " + str(round((end_time - start_time)/count, 4)) + "s) \n")
+	sys.stdout.write(" (ETR for counting kmers for hosts: " + str(round((end_time - start_time)/count*(len(hostFaList)-count), 4)) + "s) \n")
+	filelog.write(" (Average time for counting kmers for one host: " + str(round((end_time - start_time)/count, 4)) + "s) \n")
+	filelog.write(" (ETR for counting kmers for hosts: " + str(round((end_time - start_time)/count*(len(hostFaList)-count), 4)) + "s) \n")
+
+hostListFileWrite.close()
+
+
+
+#time.sleep(6) # delays for 10 seconds
+
+################### 2: compute measures #####################
+start_time = time.time()
+count = 0
+
+sys.stdout.write("Step 2: compute distance/dissimilarity measures \n")
+filelog.write("Step 2: compute distance/dissimilarity measures \n")
+cmdCptMeasure = computeMeasureOut + " -k " + str(kmax) + \
+								" -i " + virusListFile + " -j " + hostListFile + \
+								" -o " + options.outDir + " -t " + hostTaxaFile
+print(cmdCptMeasure)
+
+#with open(os.path.join(tmpDir, 'computeMeasureOut.log'), 'w') as filelog:
+cmdCptMeasureOut = subprocess.Popen(cmdCptMeasure, shell=True, \
+														 stderr = subprocess.PIPE, \
+														 stdout = subprocess.PIPE)
+#for c in iter(lambda: cmdCptMeasureOut.stderr.read(1), ''):
+for c in iter(cmdCptMeasureOut.stderr.readline, b''):
+	sys.stdout.write(c.decode("utf-8"))
+	filelog.write(c.decode("utf-8"))
+	filelog.flush()
+  
+	end_time = time.time()
+	count += 1
+	sys.stdout.write("   (Average time for computing dissimilarities for one virus-host pair: " + str(round((end_time - start_time)/count/len(virusFaList), 4)) + "s) \n")
+	sys.stdout.write("   (ETR for computing dissimilarities for virus-host pairs: " + str(round((end_time - start_time)/count*(len(hostFaList)-count), 4)) + "s) \n")
+	filelog.write("   (Average time for computing dissimilarities for one virus-host pair: " + str(round((end_time - start_time)/count/len(virusFaList), 4)) + "s) \n")
+	filelog.write("   (ETR for computing dissimilarities for virus-host pairs: " + str(round((end_time - start_time)/count*(len(hostFaList)-count), 4)) + "s) \n")
+
+cmdCptMeasureOut.wait()
+
+filelog.close()
+
+sys.stdout.write("done \n")