sequence_logo 1.0.3 → 1.0.4

data/lib/sequence_logo/ytilib/ytilib.rb

@@ -1,147 +1,147 @@
-module Ytilib
-  def Ytilib.time
-    return Time.now.strftime('%d %b %H:%M:%S')
-  end
-end
-
-$program_name = nil
-
-def start(fullpath)
-  report(fullpath + ARGV.inject("") { |out, v| out += " " + v})
-  return if $NO_REPORT
-  $program_name = "[#{File.name_wo_ext(fullpath)}]"
-end
-
-def report(message, program_name = nil)
-  $program_name = "[#{program_name}]" if program_name != nil
-  return if $NO_REPORT
-  puts "LLIB #{Ytilib.time} #{$program_name}\t#{message}" if !block_given? || yield
-end
-
-def checkerr(message = "checkerr failed")  
-  if !block_given? || yield
-    puts "LLIB #{Ytilib.time} [error]\t#{message}" unless $NO_REPORT
-    raise "LLIB #{Ytilib.time} #{$program_name}\n\t#{message}\n" 
-  end
-end
-
-module Ytilib
-  
-  STRAND_DIRECT = "direct"
-  STRAND_REVCOMP = "revcomp"  
-  
-  def Ytilib.read_mfa2hash(path)
-    input_fasta_f = File.new(path, "r")
-    seqs, seq_name = {}, nil
-    input_fasta_f.each_line { |line|
-      if line[0,1] == ">"
-        seq_name = line[1..-1].strip
-        seq_name = yield seq_name if block_given?
-        checkerr("multiple sequences with the same name=#{seq_name}") { seqs[seq_name] }
-        seqs[seq_name] = ""
-      elsif seq_name != nil
-        seqs[seq_name] << line.strip
-      end
-    }
-    input_fasta_f.close
-    return seqs
-  end
-  
-  def Ytilib.read_mfa2array(path)
-    input_fasta_f = File.new(path, "r")
-    seqs, seq_name = [], nil
-    input_fasta_f.each_line { |line|
-      if line[0,1] == ">"
-        seq_name = line[1..-1].strip
-        yield seq_name if block_given?
-        seqs << ""
-      elsif seq_name != nil
-        seqs.last << line.strip
-      end
-    }
-    input_fasta_f.close
-    return seqs
-  end
-  
-  def Ytilib.mfa2array(input)
-    seqs, seq_name = [], nil
-    input.each_line { |line|
-      if line[0,1] == ">"
-        seq_name = line[1..-1].strip
-        seqs << ""
-      elsif seq_name != nil
-        seqs.last << line.strip
-      end
-    }
-    return seqs
-  end
-  
-  def Ytilib.read_plain2array(path)
-    array = []
-    File.open(path).each_line { |line|
-      array << line.strip if !line.strip.empty?
-    }
-    return array
-  end  
-  
-  def Ytilib.read_seqs2array(path)
-    type = File.ext_wo_name(path)
-    case type
-    when "mfa", "fasta", "fa"
-      return Ytilib.read_mfa2array(path)
-    when "plain","txt"
-      return Ytilib.read_plain2array(path)
-    else
-      checkerr("unknown sequences-file, ext=#{type}")
-    end
-  end
-  
-  def Ytilib.write_mfa(seqs, path, prefix = " ")
-    if seqs.is_a?(Hash)
-      out_fasta_f = File.new(path, "w+")
-      seqs.each_key { |name|
-        out_fasta_f << ">#{prefix}#{name}" << $/ << seqs[name] << $/
-      }
-      out_fasta_f.close
-    else 
-      out_fasta_f = File.new(path, "w+")
-      seqs.each_with_index { |seq, i|
-        out_fasta_f << ">#{prefix}#{i+1}" << $/ << seq << $/
-      }
-      out_fasta_f.close
-    end
-  end
-  
-  def get_consensus(seqs)
-    report "consensus creating method should be checked, you are using unsafe code"
-    return 'nil' if seqs.size == 0
-    conslet = { 'A' => 'A', 'C' => 'C', 'G' => 'G', 'T' => 'T', 'U' => 'U',
-            'AG' => 'R', 'CT' => 'Y', 'GT' => 'K', 'AC' => 'M', 'CG' => 'S', 'AT' => 'W',
-            'CGT' => 'B', 'AGT' => 'D', 'ACT' => 'H', 'ACG' => 'V', 'ACGT' => 'N'
-    }
-    new_consensus, letters = '', []
-    0.upto(seqs[0].size-1) { |i|
-      seqs.each do |word|
-        letters << word[i] if !letters.include?(word[i])
-      end
-      letters.sort!
-      letters_string = ''
-      letters.each do |letter| letters_string << letter end
-      checkerr("cannot find consensus letter for a given letter set :#{}") { conslet[letters_string] == nil }
-      new_consensus << conslet[letters_string]	
-      letters.clear
-    }
-    return new_consensus
-  end
-
-  def Ytilib.new_mysql_conn(database)
-    my = Mysql.new(MYSQL_HOST, MYSQL_USER, MYSQL_PASSWORD, database)
-    checkerr("cannot connect to MySQL server") { my.query("select 1").fetch_row[0] != "1" }
-    return my
-  end  
-  
-
-end
-
-#report "ytilib required, working directory #{Dir.pwd}", "ytilib"
+module Ytilib
+  def Ytilib.time
+    return Time.now.strftime('%d %b %H:%M:%S')
+  end
+end
+
+$program_name = nil
+
+def start(fullpath)
+  report(fullpath + ARGV.inject("") { |out, v| out += " " + v})
+  return if $NO_REPORT
+  $program_name = "[#{File.name_wo_ext(fullpath)}]"
+end
+
+def report(message, program_name = nil)
+  $program_name = "[#{program_name}]" if program_name != nil
+  return if $NO_REPORT
+  puts "LLIB #{Ytilib.time} #{$program_name}\t#{message}" if !block_given? || yield
+end
+
+def checkerr(message = "checkerr failed")  
+  if !block_given? || yield
+    puts "LLIB #{Ytilib.time} [error]\t#{message}" unless $NO_REPORT
+    raise "LLIB #{Ytilib.time} #{$program_name}\n\t#{message}\n" 
+  end
+end
+
+module Ytilib
+  
+  STRAND_DIRECT = "direct"
+  STRAND_REVCOMP = "revcomp"  
+  
+  def Ytilib.read_mfa2hash(path)
+    input_fasta_f = File.new(path, "r")
+    seqs, seq_name = {}, nil
+    input_fasta_f.each_line { |line|
+      if line[0,1] == ">"
+        seq_name = line[1..-1].strip
+        seq_name = yield seq_name if block_given?
+        checkerr("multiple sequences with the same name=#{seq_name}") { seqs[seq_name] }
+        seqs[seq_name] = ""
+      elsif seq_name != nil
+        seqs[seq_name] << line.strip
+      end
+    }
+    input_fasta_f.close
+    return seqs
+  end
+  
+  def Ytilib.read_mfa2array(path)
+    input_fasta_f = File.new(path, "r")
+    seqs, seq_name = [], nil
+    input_fasta_f.each_line { |line|
+      if line[0,1] == ">"
+        seq_name = line[1..-1].strip
+        yield seq_name if block_given?
+        seqs << ""
+      elsif seq_name != nil
+        seqs.last << line.strip
+      end
+    }
+    input_fasta_f.close
+    return seqs
+  end
+  
+  def Ytilib.mfa2array(input)
+    seqs, seq_name = [], nil
+    input.each_line { |line|
+      if line[0,1] == ">"
+        seq_name = line[1..-1].strip
+        seqs << ""
+      elsif seq_name != nil
+        seqs.last << line.strip
+      end
+    }
+    return seqs
+  end
+  
+  def Ytilib.read_plain2array(path)
+    array = []
+    File.open(path).each_line { |line|
+      array << line.strip if !line.strip.empty?
+    }
+    return array
+  end  
+  
+  def Ytilib.read_seqs2array(path)
+    type = File.ext_wo_name(path)
+    case type
+    when "mfa", "fasta", "fa"
+      return Ytilib.read_mfa2array(path)
+    when "plain","txt"
+      return Ytilib.read_plain2array(path)
+    else
+      checkerr("unknown sequences-file, ext=#{type}")
+    end
+  end
+  
+  def Ytilib.write_mfa(seqs, path, prefix = " ")
+    if seqs.is_a?(Hash)
+      out_fasta_f = File.new(path, "w+")
+      seqs.each_key { |name|
+        out_fasta_f << ">#{prefix}#{name}" << $/ << seqs[name] << $/
+      }
+      out_fasta_f.close
+    else 
+      out_fasta_f = File.new(path, "w+")
+      seqs.each_with_index { |seq, i|
+        out_fasta_f << ">#{prefix}#{i+1}" << $/ << seq << $/
+      }
+      out_fasta_f.close
+    end
+  end
+  
+  def get_consensus(seqs)
+    report "consensus creating method should be checked, you are using unsafe code"
+    return 'nil' if seqs.size == 0
+    conslet = { 'A' => 'A', 'C' => 'C', 'G' => 'G', 'T' => 'T', 'U' => 'U',
+            'AG' => 'R', 'CT' => 'Y', 'GT' => 'K', 'AC' => 'M', 'CG' => 'S', 'AT' => 'W',
+            'CGT' => 'B', 'AGT' => 'D', 'ACT' => 'H', 'ACG' => 'V', 'ACGT' => 'N'
+    }
+    new_consensus, letters = '', []
+    0.upto(seqs[0].size-1) { |i|
+      seqs.each do |word|
+        letters << word[i] if !letters.include?(word[i])
+      end
+      letters.sort!
+      letters_string = ''
+      letters.each do |letter| letters_string << letter end
+      checkerr("cannot find consensus letter for a given letter set :#{}") { conslet[letters_string] == nil }
+      new_consensus << conslet[letters_string]	
+      letters.clear
+    }
+    return new_consensus
+  end
+
+  def Ytilib.new_mysql_conn(database)
+    my = Mysql.new(MYSQL_HOST, MYSQL_USER, MYSQL_PASSWORD, database)
+    checkerr("cannot connect to MySQL server") { my.query("select 1").fetch_row[0] != "1" }
+    return my
+  end  
+  
+
+end
+
+#report "ytilib required, working directory #{Dir.pwd}", "ytilib"
 include Ytilib

data/lib/sequence_logo/ytilib.rb

@@ -1,10 +1,10 @@
-require_relative 'ytilib/ytilib'
-require_relative 'ytilib/addon'
-require_relative 'ytilib/iupac'
-require_relative 'ytilib/pm'
-require_relative 'ytilib/pmsd'
-require_relative 'ytilib/randoom'
-require_relative 'ytilib/bismark'
-require_relative 'ytilib/hack1'
-require_relative 'ytilib/infocod'
+require_relative 'ytilib/ytilib'
+require_relative 'ytilib/addon'
+require_relative 'ytilib/iupac'
+require_relative 'ytilib/pm'
+require_relative 'ytilib/pmsd'
+require_relative 'ytilib/randoom'
+require_relative 'ytilib/bismark'
+require_relative 'ytilib/hack1'
+require_relative 'ytilib/infocod'
 require_relative 'ytilib/ppm_support'

data/lib/sequence_logo.rb

@@ -1,7 +1,7 @@
-require_relative 'sequence_logo/version'
-require_relative 'sequence_logo/pmflogo_lib'
-require_relative 'sequence_logo/cli'
-
-module SequenceLogo
-  AssetsPath = File.join(File.dirname(__FILE__), 'sequence_logo', 'assets')
-end
+require_relative 'sequence_logo/version'
+require_relative 'sequence_logo/pmflogo_lib'
+require_relative 'sequence_logo/cli'
+
+module SequenceLogo
+  AssetsPath = File.join(File.dirname(__FILE__), 'sequence_logo', 'assets')
+end

data/sequence_logo.gemspec

@@ -1,21 +1,21 @@
-# -*- encoding: utf-8 -*-
-require File.expand_path('../lib/sequence_logo/version', __FILE__)
-
-Gem::Specification.new do |gem|
-  gem.authors       = ["Ilya Vorontsov"]
-  gem.email         = ["prijutme4ty@gmail.com"]
-  gem.description   = %q{SequenceLogo is a tool for drawing sequence logos of motifs. It gets Positional Count Matrices(PCMs) or IUPAC sequences as input and generates png-logos for a motif. Also one can create logo for reverse complement or even generate logos for a whole collection of motifs.
-  Sequence logos are a graphical representation of an amino acid or nucleic acid multiple sequence alignment developed by Tom Schneider and Mike Stephens. Each logo consists of stacks of symbols, one stack for each position in the sequence. The overall height of the stack indicates the sequence conservation at that position, while the height of symbols within the stack indicates the relative frequency of each amino or nucleic acid at that position. In general, a sequence logo provides a richer and more precise description of, for example, a binding site, than would a consensus sequence (see http://weblogo.berkeley.edu/)
-}
-  gem.summary       = %q{Tool for drawing sequence logos of motifs}
-  gem.homepage      = ""
-
-  gem.files         = `git ls-files`.split($/)
-  gem.executables   = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
-  gem.test_files    = gem.files.grep(%r{^(test|spec|features)/})
-  gem.name          = "sequence_logo"
-  gem.require_paths = ["lib"]
-  gem.version       = SequenceLogo::VERSION
-  
-  gem.add_dependency('rmagick', '~> 2.13.1')
-end
+# -*- encoding: utf-8 -*-
+require File.expand_path('../lib/sequence_logo/version', __FILE__)
+
+Gem::Specification.new do |gem|
+  gem.authors       = ["Ilya Vorontsov"]
+  gem.email         = ["prijutme4ty@gmail.com"]
+  gem.description   = %q{SequenceLogo is a tool for drawing sequence logos of motifs. It gets Positional Count Matrices(PCMs) or IUPAC sequences as input and generates png-logos for a motif. Also one can create logo for reverse complement or even generate logos for a whole collection of motifs.
+  Sequence logos are a graphical representation of an amino acid or nucleic acid multiple sequence alignment developed by Tom Schneider and Mike Stephens. Each logo consists of stacks of symbols, one stack for each position in the sequence. The overall height of the stack indicates the sequence conservation at that position, while the height of symbols within the stack indicates the relative frequency of each amino or nucleic acid at that position. In general, a sequence logo provides a richer and more precise description of, for example, a binding site, than would a consensus sequence (see http://weblogo.berkeley.edu/)
+}
+  gem.summary       = %q{Tool for drawing sequence logos of motifs}
+  gem.homepage      = ""
+
+  gem.files         = `git ls-files`.split($/)
+  gem.executables   = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
+  gem.test_files    = gem.files.grep(%r{^(test|spec|features)/})
+  gem.name          = "sequence_logo"
+  gem.require_paths = ["lib"]
+  gem.version       = SequenceLogo::VERSION
+  
+  gem.add_dependency('rmagick', '~> 2.13.1')
+end

data/test/data/pcm/AHR_si.pcm

@@ -1,10 +1,10 @@
-AHR_si
-40.51343240527031  18.259112547756697  56.41253757072521  38.77363485291994
-10.877470982533044  11.870876719950774  34.66312982331297  96.54723985087516
-21.7165707818416  43.883079837598544  20.706746561638717  67.6523201955933
-2.5465132509466635  1.3171620263517245  145.8637051322628  4.231336967110781
-0.0  150.35847450464382  1.4927836298652875  2.1074592421627525
-3.441039751299748  0.7902972158110341  149.37613720253387  0.3512432070271259
-0.0  3.441039751299748  0.7024864140542533  149.81519121131782
-0.0  0.0  153.95871737667187  0.0
-43.07922333291745  66.87558226865211  16.159862546986584  27.844049228115868
+AHR_si
+40.51343240527031  18.259112547756697  56.41253757072521  38.77363485291994
+10.877470982533044  11.870876719950774  34.66312982331297  96.54723985087516
+21.7165707818416  43.883079837598544  20.706746561638717  67.6523201955933
+2.5465132509466635  1.3171620263517245  145.8637051322628  4.231336967110781
+0.0  150.35847450464382  1.4927836298652875  2.1074592421627525
+3.441039751299748  0.7902972158110341  149.37613720253387  0.3512432070271259
+0.0  3.441039751299748  0.7024864140542533  149.81519121131782
+0.0  0.0  153.95871737667187  0.0
+43.07922333291745  66.87558226865211  16.159862546986584  27.844049228115868

data/test/data/pcm/AIRE_f2.pcm

@@ -1,19 +1,19 @@
-AIRE_f2
-16.428571428571484  10.795918367346953  5.1632653061224625  8.918367346938803
-7.51020408163268  7.51020408163268  5.632653061224489  20.65306122448985
-5.632653061224489  5.632653061224489  8.448979591836776  21.591836734693906
-1.877551020408166  0.0  36.612244897959265  2.8163265306122534
-0.0  0.0  33.79591836734698  7.51020408163268
-13.14285714285717  1.877551020408166  0.938775510204083  25.346938775510242
-15.959183673469415  3.755102040816336  0.938775510204083  20.65306122448985
-15.02040816326536  5.632653061224489  5.632653061224489  15.02040816326536
-14.081632653061265  2.8163265306122534  5.632653061224489  18.775510204081698
-21.591836734693945  3.755102040816336  5.632653061224489  10.326530612244925
-15.959183673469415  1.877551020408166  2.8163265306122534  20.65306122448985
-9.38775510204083  5.632653061224489  2.8163265306122534  23.469387755102094
-0.0  0.938775510204083  36.612244897959265  3.755102040816336
-0.0  0.0  40.36734693877556  0.938775510204083
-9.38775510204083  3.755102040816336  10.326530612244925  17.836734693877606
-2.8163265306122534  7.51020408163268  7.51020408163268  23.469387755102094
-18.77551020408166  0.938775510204083  6.571428571428585  15.02040816326536
-16.663265306122497  11.030612244898006  4.45918367346938  9.153061224489816
+AIRE_f2
+16.428571428571484  10.795918367346953  5.1632653061224625  8.918367346938803
+7.51020408163268  7.51020408163268  5.632653061224489  20.65306122448985
+5.632653061224489  5.632653061224489  8.448979591836776  21.591836734693906
+1.877551020408166  0.0  36.612244897959265  2.8163265306122534
+0.0  0.0  33.79591836734698  7.51020408163268
+13.14285714285717  1.877551020408166  0.938775510204083  25.346938775510242
+15.959183673469415  3.755102040816336  0.938775510204083  20.65306122448985
+15.02040816326536  5.632653061224489  5.632653061224489  15.02040816326536
+14.081632653061265  2.8163265306122534  5.632653061224489  18.775510204081698
+21.591836734693945  3.755102040816336  5.632653061224489  10.326530612244925
+15.959183673469415  1.877551020408166  2.8163265306122534  20.65306122448985
+9.38775510204083  5.632653061224489  2.8163265306122534  23.469387755102094
+0.0  0.938775510204083  36.612244897959265  3.755102040816336
+0.0  0.0  40.36734693877556  0.938775510204083
+9.38775510204083  3.755102040816336  10.326530612244925  17.836734693877606
+2.8163265306122534  7.51020408163268  7.51020408163268  23.469387755102094
+18.77551020408166  0.938775510204083  6.571428571428585  15.02040816326536
+16.663265306122497  11.030612244898006  4.45918367346938  9.153061224489816

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: sequence_logo
 version: !ruby/object:Gem::Version
-  version: 1.0.3
+  version: 1.0.4
 platform: ruby
 authors:
 - Ilya Vorontsov
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-04-30 00:00:00.000000000 Z
+date: 2014-01-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rmagick
@@ -90,7 +90,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.0.3
+rubygems_version: 2.2.1
 signing_key: 
 specification_version: 4
 summary: Tool for drawing sequence logos of motifs
@@ -101,4 +101,3 @@ test_files:
 - test/data/logo/AIRE_f2_revcomp.png
 - test/data/pcm/AHR_si.pcm
 - test/data/pcm/AIRE_f2.pcm
-has_rdoc: