bio-mummer 0.1.0 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 8188cd9bbf60cd74e360a423ad45f805161ee649
-  data.tar.gz: 8476839ded624652d52e9bf7555fe948623351d1
+  metadata.gz: 37345c73287e9b80c15dc7426ee1045caad772f7
+  data.tar.gz: 8b57b1755f5da0ba52af5ccbb92c3d55f06964e5
 SHA512:
-  metadata.gz: 499ca02b01cf755dc9092bbe5f64eda8deb8eeb17e87eb6c666630b95b75cae5cfafb5c9bfe58dc673d59fdca7299d7ea1982ea3ac8899d1a4593aee480c87f0
-  data.tar.gz: 56922d92c400e907469a2ef73f7ea6e5b5615a1bd29413ff7011b55507d47ca15dee650ee68090246c2e7da53ee0069f7241971d4ddf8ca20d52e985680a7a8c
+  metadata.gz: c86e8bb5d957da2331d817101577aa9990e4685a6445d5b194747c069fc8d5708b93465e20ff95b4e1a37de9ab02402552c1d77a7ba1acd98ddbeb61bb98d574
+  data.tar.gz: ffb5c859d7add8bb10c9d0a8dc5dc98a51dcb9a06370633d7fe6d69a3bd8d295836bafe4e32796ba6b62a855cde9a224731aa1c1cf45e077e8cc21eb4c31fb84

data/VERSION

@@ -1 +1 @@
-0.1.0
+0.2.0

data/bin/delta2gvcf.rb

@@ -0,0 +1,181 @@
+#!/usr/bin/env ruby
+
+require 'stringio'
+require 'optparse'
+require 'ostruct'
+require 'pathname'
+require 'tmpdir'
+require 'bio-mummer'
+require 'bio'
+
+class OptParser
+  def self.parse(args)
+    options = OpenStruct.new
+
+    opt_parser = OptionParser.new do |opts|
+      opts.banner = "Usage: #{File.basename(__FILE__)} [options]"
+      opts.separator ""
+      opts.separator "Specific options:"
+
+      opts.on("-d", "--delta FILENAME", "Delta file generated by nucmer") do |fn|
+        pn = Pathname.new(fn)
+        if pn.size?
+          options.delta = pn
+          match = pn.each_line.first.match(/(?<ref>\/.*) (?<qry>\/.*)/)
+          if match
+            ref = Pathname.new(match[:ref])
+            if ref.size?
+              options.ref = Hash[Bio::FlatFile.open(ref).map{|e| [e.entry_id, e.naseq]}]
+            end
+            
+            qry = Pathname.new(match[:qry])
+            if qry.size?
+              options.qry = Hash[Bio::FlatFile.open(qry).map{|e| [e.entry_id, e.naseq]}]
+            end
+          end
+        else
+          $stderr.puts "Error: cannot read file #{fn}"
+          exit(1)
+        end
+      end
+    end
+    opt_parser.parse!(args)
+    return options
+  end
+end
+options = OptParser.parse(ARGV)
+
+# Open the Delta File
+d = BioMummer::DeltaFile.new(StringIO.new(`delta-filter -r -q #{options.delta}`))
+
+puts '##fileformat=VCFv4.1
+##ALT=<ID=NON_REF,Description="Represents any possible alternative allele at this location">
+##FILTER=<ID=LowQual,Description="Low quality">
+##FORMAT=<ID=AD,Number=.,Type=Integer,Description="Allelic depths for the ref and alt alleles in the order listed">
+##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Approximate read depth (reads with MQ=255 or with bad mates are filtered)">
+##FORMAT=<ID=GQ,Number=1,Type=Integer,Description="Genotype Quality">
+##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
+##FORMAT=<ID=MIN_DP,Number=1,Type=Integer,Description="Minimum DP observed within the GVCF block">
+##FORMAT=<ID=PL,Number=G,Type=Integer,Description="Normalized, Phred-scaled likelihoods for genotypes as defined in the VCF specification">'
+
+options.ref.each do |name, seq|
+  puts "##contig=<ID=#{name},length=#{seq.length}>"
+end
+puts "#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\tDUMMYSAMPLENAME"
+
+d.alignments
+  .group_by{|a| a.refname}.sort_by{|refname, alignments| refname}
+  .take(2)
+  .each do |refname, alignments|
+  cursor = 1
+  overlap = 0
+  alignments.sort_by{|a| [a.refstart, a.refstop].min}.each do |a|
+    ref = options.ref[a.refname].subseq(a.refstart, a.refstop)
+    qry = a.strand ? options.qry[a.queryname].subseq(a.querystart, a.querystop) : options.qry[a.queryname].subseq(a.querystart, a.querystop).complement
+    if a.refstart > cursor
+      puts [refname, cursor, '.', options.ref[refname][cursor].upcase, "<NON_REF>", '.', '.', "END=#{a.refstart - 1}", "GT:DP:GQ:MIN_DP:PL", "0:0:0:0:0,0"].join("\t")
+      cursor = a.refstart
+    elsif a.refstart < cursor
+      overlap = cursor - a.refstart
+    end
+    
+    a.distances.inject([0,0]) do |mem, distance|
+      if distance > 0
+        qry.insert(mem.first + distance - 1, ".")
+        [mem.first + distance, mem.last + distance]
+      else
+        ref.insert(mem.last - distance - 1, ".")
+        [mem.first - distance, mem.last - distance]
+      end
+    end
+    overlap2 = 0
+
+    ref.chars.zip(0..(ref.length-1), qry.chars).chunk do |refBase, i, qryBase|
+      if refBase == qryBase
+        "NOVAR"
+      elsif refBase == "."
+        "INS"
+      elsif qryBase == "."
+        "DEL"
+      else
+        "SNP"
+      end
+    end.drop_while do |varClass, arr|
+      if varClass != "INS"
+        overlap, overlap2 = overlap - arr.length , overlap
+      end
+      overlap >= 0
+    end.each do |varClass, arr|
+      case varClass
+      when "NOVAR"
+        if overlap2 > 0
+          refBase = arr[overlap2].first.upcase
+        else
+          refBase = arr.first.first.upcase
+          overlap2 = 0
+        end        
+        puts [refname,
+              cursor,
+              ".",
+              refBase,
+              "<NON_REF>",
+              '.',
+              '.',
+              "END=#{cursor + arr.length - 1 - overlap2};ALTSCAFF=#{a.queryname};RLEN=#{arr.length};OVERLAP2=#{overlap2}",
+              "GT:DP:GQ:MIN_DP:PL",
+              "0:200:200:200:0,800"].join("\t")
+        cursor += arr.length - overlap2
+        overlap2 = 0
+      when "SNP"
+        arr.each do |snp|
+        puts [refname,
+              cursor,
+              ".",
+              snp.first.upcase,
+              [snp.last.upcase, "<NON_REF>"].join(","),
+              '.',
+              '.',
+              "END=#{cursor};ALTSCAFF=#{a.queryname}",
+              "GT:DP:GQ:MIN_DP:PL",
+              "1:200:200:200:800,0,800"].join("\t")
+          cursor += 1
+        end
+        # puts [refname,
+        #       cursor,
+        #       ".",
+        #       arr.map{|a| a.first.upcase}.join,
+        #       [arr.map{|a| a.last.upcase}.join, "<NON_REF>"].join(","),
+        #       '.',
+        #       '.',
+        #       "END=#{cursor + arr.length - 1};ALTSCAFF=#{a.queryname}",
+        #       "GT:DP:GQ:MIN_DP:PL",
+        #       "1:200:200:200:800,0,800"].join("\t")
+        # cursor += arr.length
+      when "INS"
+        refBase = options.ref[refname][cursor-2].upcase
+        puts [refname,
+              cursor-1,
+              ".",
+              refBase,
+              [(refBase + arr.map{|a| a.last}.join).upcase, "<NON_REF>"].join(","),
+              '.',
+              '.',
+              "END=#{cursor - 1};ALTSCAFF=#{a.queryname};STRAND=#{a.strand}",
+              "GT:DP:GQ:MIN_DP:PL",
+              "1:200:200:200:800,0,800"].join("\t")
+      when "DEL"
+        puts [refname,
+              cursor,
+              ".",
+              arr.map{|a| a.first.upcase}.join,
+              ".,<NON_REF>",
+              '.',
+              '.',
+              "END=#{cursor + arr.length - 1};ALTSCAFF=#{a.queryname}",
+              "GT:DP:GQ:MIN_DP:PL",
+              "1:200:200:200:800,0,800"].join("\t")
+        cursor += arr.length
+      end
+    end
+  end
+end

data/lib/bio-mummer/mummer.rb

@@ -1,3 +1,4 @@
+require 'stringio'
 
 module BioMummer
 
@@ -13,25 +14,23 @@ module BioMummer
       @querystop = querystop
       @strand = strand
       @distances = distances
-      @deltas = []
     end
 
     def deltas
-      if @deltas == []
-        a = @distances.each_with_object([0]) do |d, arr|
-          state = arr.last
-          if d > 0
-            @deltas += Array.new(d - 1, state)
-            @deltas.push(nil)
-            arr << state - 1
-          else
-            @deltas += Array.new(d * -1 - 1, state)
-            arr << state + 1
-          end
+      ds = []
+      a = @distances.each_with_object([0]) do |d, arr|
+        state = arr.last
+        if d > 0
+          ds += Array.new(d - 1, state)
+          ds.push(nil)
+          arr << state - 1
+        else
+          ds += Array.new(d * -1 - 1, state)
+          arr << state + 1
         end
-        @deltas << a.last
       end
-      return @deltas
+      ds << a.last
+      return ds
     end
 
     def ref_to_query(ref_position)
@@ -58,6 +57,11 @@ module BioMummer
       @alignments = parse(io.read)
     end
 
+    def self.open(filename)
+      io = File.open(filename)
+      self.new(super(io))
+    end
+
     def parse(string)
       string.split("\n").slice_before(/^>/).flat_map do |block|
         refname, queryname = block.shift.match(/>(.*) (.*) \d+ \d+/).captures

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: bio-mummer
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
 platform: ruby
 authors:
 - robsyme
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-02-10 00:00:00.000000000 Z
+date: 2015-03-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: shoulda
@@ -97,7 +97,8 @@ dependencies:
 description: Help for working with the output of the .delta files produced by nucmer
   and promer
 email: rob.syme@gmail.com
-executables: []
+executables:
+- delta2gvcf.rb
 extensions: []
 extra_rdoc_files:
 - LICENSE.txt
@@ -112,6 +113,7 @@ files:
 - README.rdoc
 - Rakefile
 - VERSION
+- bin/delta2gvcf.rb
 - lib/bio-mummer.rb
 - lib/bio-mummer/mummer.rb
 - test/data/out.delta