miga-base 0.7.26.1 → 1.0.2.0

data/utils/enveomics/Scripts/lib/enveomics_rb/enveomics.rb

@@ -1,24 +1,24 @@
 
-#
-# @author: Luis M. Rodriguez-R
-# @license: artistic license 2.0
-#
+require 'enveomics_rb/utils'
+use 'optparse'
+ARGV << '-h' if ARGV.empty?
 
-require "optparse"
-ARGV << "-h" if ARGV.size==0
+module Enveomics
+  class << self
+    def opt_banner(opt, banner, usage = nil)
+      opt.version ||= $VERSION
+      usage ||= "#{opt.program_name}.rb [options]"
+      opt.banner = <<~BANNER
 
-def use(gems, mandatory=true)
-  gems = [gems] unless gems.is_a? Array
-  begin
-    require "rubygems"
-    while ! gems.empty?
-      require gems.shift
+        [Enveomics Collection: #{opt.program_name} #{opt.version}]
+
+        #{banner}
+
+        Usage
+          #{usage}
+
+      BANNER
     end
-    return true
-  rescue LoadError
-    abort "\nUnmet requirements, please install required gems:" +
-      gems.map{ |gem| "\n   gem install #{gem}" }.join + "\n\n" if mandatory
-    return false
   end
 end
 

data/utils/enveomics/Scripts/lib/enveomics_rb/errors.rb

@@ -0,0 +1,17 @@
+
+module Enveomics
+  class Error < RuntimeError
+  end
+
+  class CommandError < Error
+  end
+
+  class OptionError < Error
+  end
+
+  class UnimplementedError < Error
+  end
+
+  class ParseError < Error
+  end
+end

data/utils/enveomics/Scripts/lib/enveomics_rb/gmm_em.rb

@@ -0,0 +1,30 @@
+
+require 'enveomics_rb/stats/sample'
+
+module Enveomics
+  # Calculate Gaussian Mixture Models by Expectation Maximization
+  class GmmEm
+    attr :sample
+    attr :components
+    attr :opts
+
+    # Initialize Enve::GmmEm object from numeric array +x+, +components+
+    # gaussian components (an Integer), and options hash +opts+ with supported
+    # Symbol keys:
+    # - ll_delta_converge: Maximum change in LL to consider convergence
+    #   (by default: 1e-15)
+    # - max_iter: Maximum number of EM iterations (by default: 1_000)
+    # - init_mu: Initial components means as numeric array
+    # - init_sigma: Initial components standard deviation as numeric array
+    # - init_alpha: Initial components fractions as numeric array adding up to 1
+    def initialize(x, components = 2, opts = {})
+      @sample = Enve::Stats::Sample.new(x)
+      @opts = opts
+      @opts[:ll_delta_convergence] ||= 1e-15
+      @opts[:max_iter] ||= 1_000
+    end
+
+    
+  end
+end
+

data/utils/enveomics/Scripts/lib/enveomics_rb/match.rb

@@ -0,0 +1,63 @@
+
+module Enveomics
+  ##
+  # A simple object representing a sequence match from a search engine
+  # supporting tabular BLAST output
+  class Match
+    attr :row
+
+    ##
+    # Initialize Enveomics::Match object from a tabular blast line String +ln+
+    def initialize(ln)
+      @row = ln.chomp.split("\t")
+    end
+
+    def qry
+      row[0]
+    end
+
+    def sbj
+      row[1]
+    end
+
+    def id
+      @id ||= row[2].to_f
+    end
+
+    def len
+      @len ||= row[3].to_i
+    end
+
+    def evalue
+      @evalue ||= row[9].to_f
+    end
+
+    def score
+      @score ||= row[10].to_f
+    end
+
+    def qry_len
+      @qry_len ||= row[12].to_i
+    end
+
+    def sbj_len
+      @sbj_len ||= row[13].to_i
+    end
+
+    def qry_fract
+      return 0.0 unless qry_len.zero?
+      @fract ||= len.to_f / qry_len
+    end
+
+    alias fract qry_fract
+
+    def sbj_fract
+      return 0.0 unless sbj_len.zero?
+      @fract ||= len.to_f / sbj_len
+    end
+
+    def to_s
+      row.join("\t")
+    end
+  end
+end

data/utils/enveomics/Scripts/lib/enveomics_rb/rbm.rb

@@ -0,0 +1,49 @@
+require 'enveomics_rb/bm_set'
+
+module Enveomics
+  class RBM
+    attr :seq1, :seq2, :bms1, :bms2
+
+    ##
+    # Initialize RBM object with sequence paths +seq1+ and +seq2+, and
+    # Enveomics::BMset options Hash +bm_opts+
+    def initialize(seq1, seq2, bm_opts = {})
+      @seq1 = seq1
+      @seq2 = seq2
+      @bms1 = Enveomics::BMset.new(seq1, seq2, bm_opts)
+      @bms2 = Enveomics::BMset.new(seq2, seq1, bm_opts)
+      @set = nil
+    end
+
+    ##
+    # Array of Reciprocal Best Enveomics::Match objects
+    def set
+      @set ||= reciprocate!
+    end
+
+    ##
+    # Number of reciprocal best matches found
+    def count
+      set.count
+    end
+
+    ##
+    # Find reciprocal best matches and return the subset of +bms1+ that
+    # is reciprocal with +bms2+
+    def reciprocate!
+      bms1.each.select do |bm|
+        bms2[bm.sbj] && bm.qry == bms2[bm.sbj].sbj
+      end
+    end
+
+    ##
+    # Enumerate RBMs and yield +blk+
+    def each(&blk)
+      if block_given?
+        set.each { |bm| blk.call(bm) }
+      else
+        to_enum(:each)
+      end
+    end
+  end
+end

data/utils/enveomics/Scripts/lib/enveomics_rb/stats.rb

@@ -0,0 +1,3 @@
+require 'enveomics_rb/stats/rand'
+require 'enveomics_rb/stats/sample'
+

data/utils/enveomics/Scripts/lib/enveomics_rb/stats/rand.rb

@@ -0,0 +1,31 @@
+
+module Enveomics
+  module Stats
+    class << self
+      # Generates a random number from the +dist+ distribution with +params+
+      # parameters. This is simply a wrapper to the r_* functions below.
+      def rand(dist = :unif, *params)
+        send("r_#{dist}", *params)
+      end
+
+      # Generates a random number from the uniform distribution between +min+
+      # and +max+. By default generates random numbers between 0.0 and 1.0.
+      def r_unif(min = 0.0, max = 1.0)
+        min + (max - min) * Random::rand
+      end
+      
+      # Generates a random number from the geometric distribution with support
+      # {0, 1, 2, ...} and probability of success +p+.
+      def r_geom(p)
+        (Math::log(1.0 - rand) / Math::log(1.0 - p) - 1.0).ceil
+      end
+
+      # Generates a random number from the shifted geometric distribution with
+      # support {1, 2, 3, ...} and probability of success +p+.
+      def r_sgeom(p)
+        (Math::log(1.0 - rand) / Math::log(1.0 - p)).ceil
+      end
+    end
+  end
+end
+

data/utils/enveomics/Scripts/lib/enveomics_rb/stats/sample.rb

@@ -0,0 +1,152 @@
+
+module Enveomics
+  module Stats
+    # Descriptive statistics for a given sample
+    class Sample
+      attr :x
+      attr :opts
+
+      # Initialize Enveomics::Stats::Sample with numeric vector +x+ and options
+      # Hash +opts+ supporting the keys:
+      # - +effective_range+: Range where values fall (by default: range of +x+)
+      # - +histo_bin_size+: Width of histogram widths
+      #   (by default: 1/50th of +effective_range+)
+      def initialize(x, opts = {})
+        raise 'Cannot initialize an empty sample' if x.empty?
+        @x = x.map(&:to_f)
+        @opts = opts
+      end
+
+      # Size of the sample
+      def n
+        x.size
+      end
+
+      # Estimates the sample mean
+      def mean
+        @mean ||= x.inject(:+) / n
+      end
+
+      # Estimates the mean of the square of the sample
+      def square_mean
+        @square_mean ||= x.map { |i| i**2 }.inject(:+) / n
+      end
+
+      # Estimates the unbiased sample variance
+      def var
+        @var ||= (square_mean - mean ** 2) * n / (n - 1)
+      end
+
+      # Estimates the unbiased sample standard deviation
+      def sd
+        @sd ||= var ** 0.5
+      end
+
+      # --- Higher moments ---
+
+      # Estimate sample skewness
+      def skewness
+        return 0.0 if n == 1
+        cubed_dev = x.inject(0.0) { |sum, i| sum + (i - mean) ** 3 }
+        cubed_dev / ((n - 1) * (sd ** 3))
+      end
+
+      # Estimate sample excess kurtosis
+      def kurtosis
+        return 0.0 if n == 1
+        quart_dev = x.inject(0.0) { |sum, i| sum + (i - mean)**4 }
+        quart_dev / ((n - 1) * (sd**4))
+      end
+
+      # --- Ranges ---
+
+      # Range effectively considered
+      def effective_range
+        @opts[:effective_range] ||= [nil, nil]
+        @opts[:effective_range][0] ||= x.min
+        @opts[:effective_range][1] ||= x.max
+        @opts[:effective_range]
+      end
+
+      # Size of the effective range
+      def effective_range_size
+        effective_range[1] - effective_range[0]
+      end
+
+      # --- Histograms ---
+
+      # Size of each histogram bin
+      def histo_bin_size
+        @opts[:histo_bin_size] ||= effective_range_size / 50.0
+      end
+
+      # Calculate histogram ranges without checking for cached value
+      #
+      # Use #histo_ranges instead
+      def calculate_histo_ranges
+        rng = [[effective_range[1], effective_range[1] - histo_bin_size]]
+        while rng[rng.size - 1][1] > effective_range[0]
+          rng << [rng[rng.size - 1][1], rng[rng.size - 1][1] - histo_bin_size]
+        end
+        rng
+      end
+
+      # Histogram ranges as an array of two-entry arrays where the fist entry
+      # is the closed-ended maximum value (inclusive) of the range and the
+      # second entry is the open-ended minimum value (non-inclusive) of the
+      # range. The array is sorted from maximum to minimum
+      #
+      # Something like: +[[100.0, 99.0], [99.0, 98.0], ...]+, representing the
+      # ranges: {[100, 99), [99, 98), ...}
+      #
+      # The bin width is determined by #hist_bin_size
+      def histo_ranges
+        @histo_ranges ||= calculate_histo_ranges
+      end
+
+      # Mid-points of the histogram ranges from #histo_ranges, returns
+      # and array of Float
+      def histo_mids
+        @histo_mids ||= histo_ranges.map { |x| (x[0] + x[1]) / 2 }
+      end
+
+      # Calculate the histogram counts withouth checking cached value
+      #
+      # Use #histo_count instead
+      def calculate_histo_counts
+        counts = []
+        xx = x.dup
+        histo_ranges.each do |i|
+          counts << xx.size - xx.delete_if { |j| j > i[1] }.size
+        end
+        counts
+      end
+
+      # Histogram counts in the ranges determined by #histo_ranges
+      def histo_counts
+        @histo_counts ||= calculate_histo_counts
+      end
+
+      # --- Bimodality coefficients ---
+
+      # Sarle's sample bimodality coefficient b
+      def sarle_bimodality
+        (skewness**2 + 1) /
+          (kurtosis + (3 * ((n - 1)**2)) / ((n - 2) * (n - 3)))
+      end
+
+      # de Michele & Accantino (2014) B index
+      # DOI: 10.1371%2Fjournal.pone.0091195
+      def dma_bimodality
+        (mean - dma_mu_M).abs
+      end
+
+      # µ_M index proposed by Michele & Accantino (2014)
+      # DOI: 10.1371%2Fjournal.pone.0091195
+      def dma_mu_M
+        histo_counts.each_with_index.map { |m, k| m * histo_mids[k] }.inject(:+) / n
+      end
+    end
+  end
+end
+

data/utils/enveomics/Scripts/lib/enveomics_rb/utils.rb

@@ -0,0 +1,73 @@
+
+require 'enveomics_rb/errors'
+require 'zlib'
+
+def use(gems, mandatory = true)
+  gems = [gems] unless gems.is_a? Array
+  begin
+    require 'rubygems'
+    while !gems.empty?
+      require gems.shift
+    end
+    return true
+  rescue LoadError
+    abort "\nUnmet requirements, please install required gems:" +
+      gems.map{ |gem| "\n   gem install #{gem}" }.join + "\n\n" if mandatory
+    return false
+  end
+end
+
+def say(*msg)
+  return if $QUIET ||= false
+
+  o = '[%s] %s' % [Time.now, msg.join('')]
+  $stderr.puts(o)
+end
+
+##
+# Returns an open reading file handler for the file,
+# supporting .gz and '-' for STDIN
+def reader(file)
+  file == '-' ? $stdin :
+    file =~ /\.gz$/ ? Zlib::GzipReader.open(file) :
+    File.open(file, 'r')
+end
+
+##
+# Returns an open writing file handler for the file,
+# supporting .gz and '-' for STDOUT
+def writer(file)
+  file == '-' ? $stdout :
+    file =~ /\.gz$/ ? Zlib::GzipWriter.open(file) :
+    File.open(file, 'w')
+end
+
+##
+# Run a command +cmd+ that can be a ready-to-go string or an Array to escape
+# 
+# Supported symbol key options in Hash +opts+:
+# - wait: Boolean, should I wait for the command to complete? Default: true
+# - stdout: Path to redirect the standard output
+# - stderr: Path to redirect the standard error
+# - mergeout: Send stderr to stdout
+#
+# Return the process ID. If wait is true (default), check for the exit
+# status and throw an Enveomics::CommandError if non-zero
+def run_cmd(cmd, opts = {})
+  opts[:wait] = true if opts[:wait].nil?
+  cmd = cmd.shelljoin if cmd.is_a? Array
+  cmd += "  > #{opts[:stdout].shellescape}" if opts[:stdout]
+  cmd += " 2> #{opts[:stderr].shellescape}" if opts[:stderr]
+  cmd += ' 2>&1' if opts[:mergeout]
+  pid = spawn(cmd)
+  return pid unless opts[:wait]
+
+  Process.wait(pid)
+  unless $?.success?
+    raise Enveomics::CommandError.new(
+      "Command failed with status #{$?.exitstatus}:\n#{cmd}"
+    )
+  end
+  pid
+end
+