rdf-vcf 0.1.2

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 0308e08c49b3ab4dafd99a2c4d7e1eb6f290f5d8
+  data.tar.gz: 62ab5fbfbb17000b1e77ae49347cc08a43466fbd
+SHA512:
+  metadata.gz: 6c91ef56c3884e0308a5ecf9c816a531eabb2d79fb18b0ec0e6d64a1fafd2dbfb2088d83d39ba368b04ab8bbd091cb2ec4a41459e807cc6977d5b2c8fad155dc
+  data.tar.gz: 4b7ceefbdcf83a81f1aad75410c7062dd37acbecc3d6d2207faee57574908683024ed721db674b5fba882ab675f3acc8d1399564d48a7352f6b7b36b5b97920d

data/AUTHORS

@@ -0,0 +1,3 @@
+* Arto Bendiken <arto@bendiken.net>
+* Raoul J.P. Bonnal <ilpuccio.febo@gmail.com>
+* Francesco Strozzi <francesco.strozzi@gmail.com>

data/README

@@ -0,0 +1,79 @@
+RDF::VCF
+========
+
+This is an [RDF.rb](https://github.com/ruby-rdf/rdf) reader plugin for
+Variant Call Format (VCF) files, widely used in bioinformatics.
+
+This project grew out of [BioHackathon 2014](http://2014.biohackathon.org/)
+[work](https://github.com/dbcls/bh14/wiki/On-The-Fly-RDF-converter) by
+[Raoul J.P. Bonnal](https://github.com/helios) and [Francesco
+Strozzi](https://github.com/fstrozzi), and was further developed during
+[BioHackathon 2015](http://2015.biohackathon.org/).
+
+Note: at present, the project requires JRuby 9.0 (or newer) due to the
+Java-based VCF parser. We hope to eventually substitute the pure-Ruby
+[Bio-vcf](https://github.com/pjotrp/bioruby-vcf) instead.
+
+Features
+========
+
+* Implements an RDF.rb reader for VCF and BCF files, supporting also
+  bgzipped files.
+* Includes a CLI tool called `vcf2rdf` to transform VCF files into RDF.
+
+Examples
+========
+
+Reading VCF Files
+-----------------
+
+The gem can be used like any other RDF.rb reader plugin:
+
+    require 'rdf/vcf'
+
+    RDF::VCF::Reader.open('Homo_sapiens.vcf.gz') do |reader|
+      reader.each_statement do |statement|
+        p statement
+      end
+    end
+
+Command-Line Interface (CLI)
+----------------------------
+
+The gem includes a CLI tool called `vcf2rdf` which transforms VCF files into
+RDF (currently, N-Triples):
+
+    vcf2rdf your_vcf_file.vcf.gz
+
+Input files can be either plain text VCF or else compressed by `bgzip`, as
+above.
+
+Notes
+-----
+
+Please create the [`tabix`](http://www.htslib.org/doc/tabix.html) index by
+youself. In the future, this gem will create the index automatically.
+
+Dependencies
+============
+
+* [JRuby](http://jruby.org) (>= 9.0)
+* [RDF.rb](https://github.com/ruby-rdf/rdf) (>= 1.1)
+
+Mailing List
+============
+
+* http://lists.w3.org/Archives/Public/public-rdf-ruby/
+
+Authors
+=======
+
+* [Arto Bendiken](https://github.com/bendiken)
+* [Raoul J.P. Bonnal](https://github.com/helios)
+* [Francesco Strozzi](https://github.com/fstrozzi)
+
+License
+=======
+
+This is free and unencumbered public domain software. For more information,
+see <http://unlicense.org/> or the accompanying {file:UNLICENSE} file.

data/UNLICENSE

@@ -0,0 +1,24 @@
+This is free and unencumbered software released into the public domain.
+
+Anyone is free to copy, modify, publish, use, compile, sell, or
+distribute this software, either in source code form or as a compiled
+binary, for any purpose, commercial or non-commercial, and by any
+means.
+
+In jurisdictions that recognize copyright laws, the author or authors
+of this software dedicate any and all copyright interest in the
+software to the public domain. We make this dedication for the benefit
+of the public at large and to the detriment of our heirs and
+successors. We intend this dedication to be an overt act of
+relinquishment in perpetuity of all present and future rights to this
+software under copyright law.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
+OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+OTHER DEALINGS IN THE SOFTWARE.
+
+For more information, please refer to <http://unlicense.org/>

data/VERSION

@@ -0,0 +1 @@
+0.1.2

data/bin/vcf2rdf

@@ -0,0 +1,23 @@
+#!/usr/bin/env jruby
+$:.unshift(File.expand_path('../lib', __dir__))
+require 'rdf/vcf'
+
+if ARGV.empty? || ARGV.include?('--help') || ARGV.include?('-h')
+  abort "usage: #{$0} file.vcf[.gz]..."
+end
+
+ARGV.each do |pathname|
+  # TODO: support an --output=MIME option:
+  writer = RDF::Writer.for(:nquads).new($stdout)
+
+  RDF::VCF::Reader.open(pathname) do |reader|
+    reader.each_statement do |statement|
+      begin
+        writer << statement
+      rescue => error
+        # @see https://github.com/ruby-rdf/rdf/issues/210
+        warn error  # skip statements that cannot be serialized
+      end
+    end
+  end
+end

data/lib/rdf-vcf.rb

@@ -0,0 +1 @@
+require 'rdf/vcf'

data/lib/rdf/vcf.rb

@@ -0,0 +1,3 @@
+require 'rdf/vcf/reader'
+require 'rdf/vcf/record'
+require 'rdf/vcf/version'

data/lib/rdf/vcf/reader.rb

@@ -0,0 +1,146 @@
+require 'java' # requires JRuby
+require 'rdf/vcf/jar/htsjdk-1.139.jar'
+require 'rdf/vcf/jar/bzip2.jar'
+require 'digest/md5'
+
+require 'rdf/vcf/record'
+
+module RDF; module VCF
+  ##
+  # VCF file reader.
+  #
+  # This is a user-friendly wrapper for the HTSJDK implementation.
+  #
+  # @see https://github.com/samtools/htsjdk
+  # @see https://samtools.github.io/htsjdk/javadoc/htsjdk/htsjdk/variant/vcf/VCFFileReader.html
+  class Reader
+    java_import 'htsjdk.variant.vcf.VCFFileReader'
+
+    REF_BASE_URI = 'http://rdf.ebi.ac.uk/resource/ensembl/%s/chromosome:%s:%s'.freeze
+
+    def ref_base_uri #
+      @file_iri ||= RDF::URI(REF_BASE_URI % [reference_md5, file_date || "" , source || ""])
+    end
+
+
+    ##
+    # @param [#to_s] pathname
+    def self.open(pathname, &block)
+      reader = self.new(pathname)
+      block.call(reader)
+    ensure
+      #reader.close
+    end
+
+    ##
+    # @param [#to_s] pathname
+    def initialize(pathname)
+      pathname = pathname.to_s
+      @vcf_file = java.io.File.new(pathname)
+      @tbi_file = java.io.File.new("#{pathname}.tbi") rescue nil
+      @reader = VCFFileReader.new(@vcf_file, @tbi_file, false)
+    end
+
+    ##
+    # @return [Boolean]
+    def closed?
+      @reader.nil?
+    end
+
+    ##
+    # @return [void]
+    def close
+      @reader.close if @reader
+    ensure
+      @reader, @vcf_file, @tbi_file = nil, nil, nil
+    end
+
+    ##
+    # @yield  [statement]
+    # @yieldparam  [RDF::Statement] statement
+    # @yieldreturn [void]
+    # @return [void]
+    def each_statement(&block)
+      self.each_record do |record|
+        record.to_rdf.each(&block)
+      end
+    end
+
+    ##
+    # @yield  [record]
+    # @yieldparam  [Record] record
+    # @yieldreturn [void]
+    # @return [void]
+    def each_record(&block)
+      return unless @reader
+      @reader.iterator.each do |variant_context| # VariantContext
+        record = Record.new(variant_context, self)
+        block.call(record)
+      end
+    end
+
+    ##
+    # @param  [String]  chromosome
+    # @param  [Integer] start_pos
+    # @param  [Integer] end_pos
+    # @yield  [record]
+    # @yieldparam  [Record] record
+    # @yieldreturn [void]
+    # @return [void]
+    def find_records(chromosome: nil, start_pos: nil, end_pos: nil, &block)
+      return unless @reader
+      start_pos  ||= 0
+      end_pos    ||= java.lang.Integer::MAX_VALUE
+      @reader.query(chromosome, start_pos, end_pos).each do |variant_context| # VariantContext
+        record = Record.new(variant_context, self)
+        block.call(record)
+      end
+    end
+
+    ##
+    # @param  [Integer] pos
+    # @return [Boolean]
+    def has_position?(pos)
+      true # TODO
+    end
+
+    ##
+    # @return [String]
+    def reference()
+      @reference ||= @reader.getFileHeader().getMetaDataLine("reference").getKey rescue nil
+    end
+
+    def reference_md5
+      ::Digest::MD5.hexdigest(reference|| "")
+    end
+
+
+    ##
+    # @return [String]
+    def file_date
+      @file_date ||= @reader.getFileHeader().getMetaDataLine("fileDate").getKey rescue nil
+    end
+
+    ##
+    # @return [String]
+    def source
+      @source ||= @reader.getFileHeader().getMetaDataLine("source").getKey rescue nil
+    end
+
+
+
+  end # Reader
+end; end # RDF::VCF
+
+if $0 == __FILE__
+  VCF::Reader.open('Homo_sapiens.vcf.gz') do |file|
+    p file
+    file.find_records(chromosome: "Y") do |record|
+    #file.each_record do |record|
+      p record
+      #record.to_rdf.each do |statement|
+      #  p statement
+      #end
+    end
+  end
+end

data/lib/rdf/vcf/record.rb

@@ -0,0 +1,168 @@
+require 'java' # requires JRuby
+require 'rdf/vcf/jar/htsjdk-1.139.jar'
+require 'rdf/vcf/jar/bzip2.jar'
+
+require 'digest/md5'
+require 'rdf'
+
+module RDF; module VCF
+  ##
+  # VCF file record.
+  #
+  # This is a user-friendly wrapper for the HTSJDK implementation.
+  #
+  # @see https://github.com/samtools/htsjdk
+  # @see https://samtools.github.io/htsjdk/javadoc/htsjdk/htsjdk/variant/variantcontext/VariantContext.html
+  class Record
+    include RDF
+    java_import 'htsjdk.variant.variantcontext.VariantContext'
+
+    FALDO = RDF::Vocabulary.new("http://biohackathon.org/resource/faldo#")
+
+    VAR_BASE_URI = 'http://rdf.ebi.ac.uk/terms/ensemblvariation/%s'.freeze
+
+    ##
+    # @param [VariantContext] variant_context
+    def initialize(variant_context, reader)
+      @vcf = variant_context
+      @reader = reader
+    end
+
+    ##
+    # @return [String]
+    def id
+      @id ||= case (id = @vcf.getID)
+        when '.'
+          label = "%s:%s:%s-%s" % ['', @vcf.getChr, @vcf.getStart, @vcf.getEnd] # TODO: species
+          ::Digest::MD5.hexdigest(label)
+        else id
+      end
+    end
+
+    ##
+    # @return [String]
+    def uri
+      @uri ||= VAR_BASE_URI % self.id
+    end
+
+    ##
+    # @return [String]
+    def chromosome
+      @vcf.getChr
+    end
+
+    def start
+      @vcf.getStart
+    end
+
+    def stop
+      @vcf.getEnd
+    end
+
+    ##
+    # @return [Hash{String => String}]
+    def attributes
+      @vcf.getAttributes
+    end
+
+    ##
+    # @return [String]
+    def reference
+      @reader.reference
+    end
+
+    ##
+    # @return [String]
+    def file_date
+      @reader.file_date
+    end
+
+    ##
+    # @return [String]
+    def source
+      @reader.source
+    end
+
+    ##
+    # @return [URI]
+    def ref_base_uri
+      @reader.ref_base_uri
+    end
+
+
+    def get_alleles(&block)
+      @vcf.getAlleles.each do |allele|
+        block.call(allele)
+      end
+    end
+
+    def get_alternate_alleles(&block)
+      @vcf.getAlternateAlleles.each do |allele|
+        block.call(allele)
+      end
+    end
+    ##
+    # @return [String]
+    def get_reference_allele
+      @vcf.getReference.getBaseString
+    end
+
+    ##
+    # @return [String]
+    # def get_alternate_allele
+    #   @vcf.getAlternateAlleles.first.getBaseString
+    # end
+
+    ##
+    # @return [Double]
+    def get_phred_scaled_qual
+      @vcf.getPhredScaledQual
+    end
+
+    ##
+    # @return [RDF::Graph]
+    def to_rdf
+      var_uri = RDF::URI(self.uri)
+      RDF::Graph.new do |graph|
+        graph << [var_uri, RDF::DC.identifier, self.id]
+        graph << [var_uri, RDF::RDFS.label, self.id]
+        graph << [self.ref_base_uri,DC.identifier,self.id]
+        faldoRegion = ref_base_uri+":#{self.start}-#{self.stop}:1"
+        graph << [var_uri,FALDO.location,faldoRegion]
+        graph << [faldoRegion,RDFS.label,"#{self.id}:#{self.start}-#{self.stop}:1"]
+        graph << [faldoRegion,RDF.type,FALDO.Region]
+        graph << [faldoRegion,FALDO.begin,self.ref_base_uri+":#{self.start}:1"]
+        graph << [faldoRegion,FALDO.end,self.ref_base_uri+":#{self.stop}:1"]
+        graph << [faldoRegion,FALDO.reference,self.ref_base_uri]
+        if self.start == self.stop
+          faldoExactPosition = self.ref_base_uri+":#{self.start}:1"
+          graph << [faldoExactPosition,RDF.type,"faldo:ExactPosition"]
+          graph << [faldoExactPosition,RDF.type,"faldo:ForwardStrandPosition"]
+          graph << [faldoExactPosition,FALDO.position,self.start]
+          graph << [faldoExactPosition,FALDO.reference,self.ref_base_uri]
+        end
+
+        # TODO check if there are multiple alleles and iterate over them
+        # TODO what happends if there is an insertion?
+        refAlleleURI = var_uri+"\##{get_reference_allele}"
+        graph << [var_uri,var_uri+":has_allele",refAlleleURI]
+        graph << [refAlleleURI,RDFS.label,"#{self.id} allele #{get_reference_allele}"]
+        graph << [refAlleleURI,RDF.type,var_uri+":reference_allele"]
+        self.get_alternate_alleles do |altAllele|
+          altAlleleURI = var_uri+"\##{altAllele.getBaseString}"
+          graph << [var_uri,var_uri+":has_allele",altAlleleURI]
+          graph << [altAlleleURI,RDFS.label,"#{self.id} allele #{altAllele.getBaseString}"]
+          graph << [altAlleleURI, RDF.type, var_uri+":ancestral_allele"]
+
+        end
+        graph << [var_uri,URI(VAR_BASE_URI % "/vcf/quality"), RDF::Literal(get_phred_scaled_qual)]
+
+        @vcf.attributes.each do |k, v|
+          graph << [var_uri, RDF::URI( VAR_BASE_URI % "vcf/attribute\##{k}"), v] if v
+        end
+
+      end #new graph
+
+    end #to_rdf
+  end # Record
+end; end # RDF::VCF

data/lib/rdf/vcf/version.rb

@@ -0,0 +1,19 @@
+module RDF; module VCF
+  module VERSION
+    FILE = File.expand_path('../../../VERSION', __dir__).freeze
+    STRING = File.read(FILE).chomp.freeze
+    MAJOR, MINOR, TINY, EXTRA = STRING.split('.').map(&:to_i).freeze
+
+    ##
+    # @return [String]
+    def self.to_s() STRING end
+
+    ##
+    # @return [String]
+    def self.to_str() STRING end
+
+    ##
+    # @return [Array(Integer, Integer, Integer)]
+    def self.to_a() [MAJOR, MINOR, TINY].freeze end
+  end # VERSION
+end; end # RDF::VCF

metadata

@@ -0,0 +1,119 @@
+--- !ruby/object:Gem::Specification
+name: rdf-vcf
+version: !ruby/object:Gem::Version
+  version: 0.1.2
+platform: ruby
+authors:
+- Arto Bendiken
+- Raoul J.P. Bonnal
+- Francesco Strozzi
+autorequire:
+bindir: bin
+cert_chain: []
+date: 2015-09-18 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.1'
+  name: rdf
+  prerelease: false
+  type: :runtime
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.1'
+- !ruby/object:Gem::Dependency
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '10'
+  name: rake
+  prerelease: false
+  type: :development
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '10'
+- !ruby/object:Gem::Dependency
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3'
+  name: rspec
+  prerelease: false
+  type: :development
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3'
+- !ruby/object:Gem::Dependency
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.8'
+  name: yard
+  prerelease: false
+  type: :development
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.8'
+description: |2
+      This is an RDF.rb reader plugin for Variant Call Format (VCF) files,
+      widely used in bioinformatics.
+email: public-rdf-ruby@w3.org
+executables:
+- vcf2rdf
+extensions: []
+extra_rdoc_files: []
+files:
+- AUTHORS
+- CREDITS
+- README
+- UNLICENSE
+- VERSION
+- bin/vcf2rdf
+- lib/rdf-vcf.rb
+- lib/rdf/vcf.rb
+- lib/rdf/vcf/jar/bzip2.jar
+- lib/rdf/vcf/jar/htsjdk-1.139.jar
+- lib/rdf/vcf/reader.rb
+- lib/rdf/vcf/record.rb
+- lib/rdf/vcf/version.rb
+homepage: https://github.com/ruby-rdf/rdf-vcf
+licenses:
+- Public Domain
+metadata: {}
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '2.2'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '2.2'
+requirements:
+- JRuby (>= 9.0)
+rubyforge_project:
+rubygems_version: 2.4.8
+signing_key:
+specification_version: 4
+summary: RDF.rb reader for Variant Call Format (VCF) files.
+test_files: []
+has_rdoc: false