X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=code_ruby%2Flib%2Fmaasha%2Fseq.rb;h=e4b6bfd7c27d5933d0bd9c6b65210bd9a7b32ea0;hb=b2ea0b5a51a558478af60e9df4c643dd58552086;hp=f2d93d1fc5bbf86b6611e56e9eb9ce98d9185795;hpb=d4b9c1bec954d5ff3e6eb2f7dc6c1aaa1f7f810f;p=biopieces.git diff --git a/code_ruby/lib/maasha/seq.rb b/code_ruby/lib/maasha/seq.rb index f2d93d1..e4b6bfd 100644 --- a/code_ruby/lib/maasha/seq.rb +++ b/code_ruby/lib/maasha/seq.rb @@ -27,8 +27,10 @@ require 'maasha/seq/digest' require 'maasha/seq/trim' require 'narray' -autoload :BackTrack, 'maasha/seq/backtrack.rb' -autoload :Dynamic, 'maasha/seq/dynamic.rb' +autoload :BackTrack, 'maasha/seq/backtrack' +autoload :Dynamic, 'maasha/seq/dynamic' +autoload :Homopolymer, 'maasha/seq/homopolymer' +autoload :Levenshtein, 'maasha/seq/levenshtein' # Residue alphabets DNA = %w[a t c g] @@ -67,15 +69,15 @@ TRANS_TAB11 = { "GTG" => "V", "GCG" => "A", "GAG" => "E", "GGG" => "G" } -# Quality scores bases -SCORE_BASE = 64 -SCORE_MIN = 0 -SCORE_MAX = 40 - # Error class for all exceptions to do with Seq. class SeqError < StandardError; end class Seq + # Quality scores bases + SCORE_BASE = 33 + SCORE_MIN = 0 + SCORE_MAX = 40 + include Digest include Trim @@ -86,7 +88,7 @@ class Seq def self.new_bp(record) seq_name = record[:SEQ_NAME] seq = record[:SEQ] - type = record[:SEQ_TYPE] + type = record[:SEQ_TYPE].to_sym if record[:SEQ_TYPE] qual = record[:SCORES] self.new(seq_name, seq, type, qual) @@ -97,9 +99,9 @@ class Seq raise SeqError, "Cannot generate negative oligo length: #{length}" if length <= 0 case type.downcase - when /dna/ then alph = DNA - when /rna/ then alph = RNA - when /protein/ then alph = PROTEIN + when :dna then alph = DNA + when :rna then alph = RNA + when :protein then alph = PROTEIN else raise SeqError, "Unknown sequence type: #{type}" end @@ -139,9 +141,9 @@ class Seq raise SeqError, "Guess failed: sequence is nil" if self.seq.nil? case self.seq[0 ... 100].downcase - when /[flpqie]/ then return "protein" - when /[u]/ then return "rna" - else return "dna" + when /[flpqie]/ then return :protein + when /[u]/ then return :rna + else return :dna end end @@ -149,6 +151,7 @@ class Seq # by inspecting the first 100 residues. def type_guess! self.type = self.type_guess + self end # Returns the length of a sequence. @@ -188,24 +191,24 @@ class Seq # Method that returns true is a given sequence type is DNA. def is_dna? - self.type == 'dna' + self.type == :dna end # Method that returns true is a given sequence type is RNA. def is_rna? - self.type == 'rna' + self.type == :rna end # Method that returns true is a given sequence type is protein. def is_protein? - self.type == 'protein' + self.type == :protein end # Method to transcribe DNA to RNA. def to_rna raise SeqError, "Cannot transcribe 0 length sequence" if self.length == 0 raise SeqError, "Cannot transcribe sequence type: #{self.type}" unless self.is_dna? - self.type = 'rna' + self.type = :rna self.seq.tr!('Tt','Uu') end @@ -213,14 +216,13 @@ class Seq def to_dna raise SeqError, "Cannot reverse-transcribe 0 length sequence" if self.length == 0 raise SeqError, "Cannot reverse-transcribe sequence type: #{self.type}" unless self.is_rna? - - self.type = 'dna' + self.type = :dna self.seq.tr!('Uu','Tt') end # Method to translate a DNA sequence to protein. def translate!(trans_tab = 11) - raise SeqError, "Sequence type must be 'dna' - not #{self.type}" unless self.type == 'dna' + raise SeqError, "Sequence type must be 'dna' - not #{self.type}" unless self.type == :dna raise SeqError, "Sequence length must be a multiplum of 3 - was: #{self.length}" unless (self.length % 3) == 0 case trans_tab @@ -255,7 +257,7 @@ class Seq self.seq = protein self.qual = nil - self.type = "protein" + self.type = :protein self end @@ -333,17 +335,13 @@ class Seq key end - # Method to reverse complement sequence. - def reverse_complement - self.reverse - self.complement - self + # Method to reverse the sequence. + def reverse + Seq.new(self.seq_name, self.seq.reverse, self.type, self.qual ? self.qual.reverse : self.qual) end - alias :revcomp :reverse_complement - # Method to reverse the sequence. - def reverse + def reverse! self.seq.reverse! self.qual.reverse! if self.qual self @@ -353,6 +351,26 @@ class Seq def complement raise SeqError, "Cannot complement 0 length sequence" if self.length == 0 + entry = Seq.new + entry.seq_name = self.seq_name + entry.type = self.type + entry.qual = self.qual + + if self.is_dna? + entry.seq = self.seq.tr('AGCUTRYWSMKHDVBNagcutrywsmkhdvbn', 'TCGAAYRWSKMDHBVNtcgaayrwskmdhbvn') + elsif self.is_rna? + entry.seq = self.seq.tr('AGCUTRYWSMKHDVBNagcutrywsmkhdvbn', 'UCGAAYRWSKMDHBVNucgaayrwskmdhbvn') + else + raise SeqError, "Cannot complement sequence type: #{self.type}" + end + + entry + end + + # Method that complements sequence including ambiguity codes. + def complement! + raise SeqError, "Cannot complement 0 length sequence" if self.length == 0 + if self.is_dna? self.seq.tr!('AGCUTRYWSMKHDVBNagcutrywsmkhdvbn', 'TCGAAYRWSKMDHBVNtcgaayrwskmdhbvn') elsif self.is_rna? @@ -360,41 +378,90 @@ class Seq else raise SeqError, "Cannot complement sequence type: #{self.type}" end + + self end # Method to determine the Hamming Distance between # two Sequence objects (case insensitive). - def hamming_distance(seq) - self.seq.upcase.hamming_distance(seq.seq.upcase) + def hamming_distance(entry) + self.seq.upcase.hamming_distance(entry.seq.upcase) + end + + # Method to determine the Edit Distance between + # two Sequence objects (case insensitive). + def edit_distance(entry) + Levenshtein.distance(self.seq, entry.seq) end # Method that generates a random sequence of a given length and type. def generate(length, type) raise SeqError, "Cannot generate sequence length < 1: #{length}" if length <= 0 - case type.downcase - when "dna" - alph = DNA - when "rna" - alph = RNA - when "protein" - alph = PROTEIN + case type + when :dna then alph = DNA + when :rna then alph = RNA + when :protein then alph = PROTEIN else raise SeqError, "Unknown sequence type: #{type}" end seq_new = Array.new(length) { alph[rand(alph.size)] }.join("") self.seq = seq_new - self.type = type.downcase + self.type = type seq_new end - # Method to shuffle a sequence readomly inline. + # Method to return a new Seq object with shuffled sequence. + def shuffle + Seq.new(self.seq_name, self.seq.split('').shuffle!.join, self.type, self.qual) + end + + # Method to shuffle a sequence randomly inline. def shuffle! self.seq = self.seq.split('').shuffle!.join self end + # Method to add two Seq objects. + def +(entry) + new_entry = Seq.new() + new_entry.seq = self.seq + entry.seq + new_entry.type = self.type if self.type == entry.type + new_entry.qual = self.qual + entry.qual if self.qual and entry.qual + new_entry + end + + # Method to concatenate sequence entries. + def <<(entry) + raise SeqError, "sequences of different types" unless self.type == entry.type + raise SeqError, "qual is missing in one entry" unless self.qual.class == entry.qual.class + + self.seq << entry.seq + self.qual << entry.qual unless entry.qual.nil? + + self + end + + # Index method for Seq objects. + def [](*args) + entry = Seq.new + entry.seq_name = self.seq_name + entry.seq = self.seq[*args] + entry.type = self.type + entry.qual = self.qual[*args] unless self.qual.nil? + + entry + end + + # Index assignment method for Seq objects. + def []=(*args, entry) + self.seq[*args] = entry.seq[*args] + self.qual[*args] = entry.qual[*args] unless self.qual.nil? + + self + end + # Method that returns a subsequence of from a given start position # and of a given length. def subseq(start, length = self.length - start) @@ -412,7 +479,9 @@ class Seq qual = self.qual[start .. stop] unless self.qual.nil? end - Seq.new(self.seq_name, seq, self.type, qual) # TODO changed self.seq_name.dup to self.seq_name -> consequence? + seq_name = self.seq_name.nil? ? nil : self.seq_name.dup + + Seq.new(seq_name, seq, self.type, qual) end # Method that replaces a sequence with a subsequence from a given start position @@ -453,23 +522,6 @@ class Seq comp end - # Method that returns the length of the longest homopolymeric stretch - # found in a sequence. - def homopol_max(min = 1) - return 0 if self.seq.nil? or self.seq.empty? - - found = false - - self.seq.upcase.scan(/A{#{min},}|T{#{min},}|G{#{min},}|C{#{min},}|N{#{min},}/) do |match| - found = true - min = match.size > min ? match.size : min - end - - return 0 unless found - - min - end - # Method that returns the percentage of hard masked residues # or N's in a sequence. def hard_mask @@ -482,21 +534,6 @@ class Seq ((self.seq.scan(/[a-z]/).size.to_f / (self.len - self.indels).to_f) * 100).round(2) end - # Hard masks sequence residues where the corresponding quality score - # is below a given cutoff. - def mask_seq_hard_old(cutoff) - seq = self.seq.upcase - scores = self.qual - i = 0 - - scores.each_char do |score| - seq[i] = 'N' if score.ord - SCORE_BASE < cutoff - i += 1 - end - - self.seq = seq - end - # Hard masks sequence residues where the corresponding quality score # is below a given cutoff. def mask_seq_hard!(cutoff) @@ -540,61 +577,51 @@ class Seq def qual_base33? self.qual.match(/[!-:]/) ? true : false end - - # Method that determines if a quality score string can be - # absolutely identified as base 64. + + # Method that determines if a quality score string may be base 64. def qual_base64? self.qual.match(/[K-h]/) ? true : false end - # Method to determine if a quality score is valid. + # Method to determine if a quality score is valid accepting only 0-40 range. def qual_valid?(encoding) raise SeqError, "Missing qual" if self.qual.nil? - case encoding.downcase - when "sanger" then return true if self.qual.match(/^[!-I]*$/) - when "454" then return true if self.qual.match(/^[@-h]*$/) - when "solexa" then return true if self.qual.match(/^[;-h]*$/) - when "illumina13" then return true if self.qual.match(/^[@-h]*$/) - when "illumina15" then return true if self.qual.match(/^[@-h]*$/) - when "illumina18" then return true if self.qual.match(/^[!-J]*$/) + case encoding + when :base_33 then return true if self.qual.match(/^[!-I]*$/) + when :base_64 then return true if self.qual.match(/^[@-h]*$/) else raise SeqError, "unknown quality score encoding: #{encoding}" end false end - # Method to convert quality scores inbetween formats. - # Sanger base 33, range 0-40 - # 454 base 64, range 0-40 - # Solexa base 64, range -5-40 - # Illumina13 base 64, range 0-40 - # Illumina15 base 64, range 0-40 - # Illumina18 base 33, range 0-41 - def convert_scores!(from, to) - unless from == to - na_qual = NArray.to_na(self.qual, "byte") + # Method to coerce quality scores to be within the 0-40 range. + def qual_coerce!(encoding) + raise SeqError, "Missing qual" if self.qual.nil? - case from.downcase - when "sanger" then na_qual -= 33 - when "454" then na_qual -= 64 - when "solexa" then na_qual -= 64 - when "illumina13" then na_qual -= 64 - when "illumina15" then na_qual -= 64 - when "illumina18" then na_qual -= 33 - else raise SeqError, "unknown quality score encoding: #{from}" - end + case encoding + when :base_33 then self.qual.tr!("[J-~]", "I") + when :base_64 then self.qual.tr!("[i-~]", "h") + else raise SeqError, "unknown quality score encoding: #{encoding}" + end - case to.downcase - when "sanger" then na_qual += 33 - when "454" then na_qual += 64 - when "solexa" then na_qual += 64 - when "illumina13" then na_qual += 64 - when "illumina15" then na_qual += 64 - when "illumina18" then na_qual += 33 - else raise SeqError, "unknown quality score encoding: #{to}" - end + self + end + + # Method to convert quality scores. + def qual_convert!(from, to) + raise SeqError, "unknown quality score encoding: #{from}" unless from == :base_33 or from == :base_64 + raise SeqError, "unknown quality score encoding: #{to}" unless to == :base_33 or to == :base_64 + if from == :base_33 and to == :base_64 + na_qual = NArray.to_na(self.qual, "byte") + na_qual += 64 - 33 + self.qual = na_qual.to_s + elsif from == :base_64 and to == :base_33 + self.qual.tr!("[;-?]", "@") # Handle negative Solexa values from -5 to -1 (set these to 0). + na_qual = NArray.to_na(self.qual, "byte") + na_qual -= 64 - 33 self.qual = na_qual.to_s end @@ -607,6 +634,7 @@ class Seq na_qual = NArray.to_na(self.qual, "byte") na_qual -= SCORE_BASE + na_qual.mean end