bacterial-annotator 0.0.1

data/lib/bacterial-annotator/fasta-manip.rb

@@ -0,0 +1,166 @@
+# -*- coding: utf-8 -*-
+# author:  	maxime déraspe
+# email:	maxime@deraspe.net
+# review:  	
+# date:    	15-02-24
+# version: 	0.0.1
+# licence:  	
+
+
+
+class FastaManip
+
+  attr_reader :fasta_flat, :fasta_file, :prodigal_files
+
+  # Initialize fasta holder
+  def initialize fasta_file, meta
+
+    @fasta_file = fasta_file
+    @fasta_flat = Bio::FlatFile.auto(@fasta_file)
+    @meta = meta
+    @prodigal_files = nil
+    @single_fasta = nil
+
+    if @fasta_flat.dbclass != Bio::FastaFormat
+      abort "Aborting : The input sequence is not a fasta file !"
+    end
+
+  end
+
+  # Run prodigal on the genome to annotate
+  def run_prodigal root, outdir
+    @prodigal_files = {}
+    Dir.mkdir "#{outdir}" if ! Dir.exists? "#{outdir}"
+    if @meta
+      system("#{root}/prodigal.linux -p meta -i #{@fasta_file} -a #{outdir}/Proteins.fa -d #{outdir}/Genes.fa -o #{outdir}/Genbanks.gbk -q")
+    else
+      system("#{root}/prodigal.linux -i #{@fasta_file} -a #{outdir}/Proteins.fa -d #{outdir}/Genes.fa -o #{outdir}/Genbanks.gbk -q")
+    end
+
+    @prodigal_files = {multiGBK: "#{outdir}/Genbanks.gbk",
+                       contigs: [],
+                       contigs_length: [],
+                       genes: "#{outdir}/Genes.fa",
+                       proteins: "#{outdir}/Proteins.fa",
+                       prot_ids_by_contig: {},
+                       fasta_path: "#{outdir}/single-fasta/",
+                       gbk_path: "#{outdir}/single-genbank/"}
+    split_fasta outdir
+    split_genbank outdir, "#{outdir}/Genbanks.gbk"
+    extract_cds_names
+    @prodigal_files
+  end
+
+
+  # Split Multi Genbanks file
+  # RETURN : array of fasta files
+  def split_fasta outdir
+    @single_fasta = {}
+    Dir.mkdir("#{outdir}/single-fasta") if ! Dir.exists?("#{outdir}/single-fasta")
+    @fasta_flat.each_entry do |seq|
+      file_name = seq.definition.chomp.split(" ")[0]
+      @prodigal_files[:contigs] << "#{file_name}"
+      @prodigal_files[:contigs_length] << seq.seq.length
+      File.open("#{outdir}/single-fasta/#{file_name}.fasta", "w") do |fwrite| 
+        fwrite.write(seq)
+      end
+      @single_fasta[file_name] = seq
+    end
+  end
+
+
+  # Split Multi Genbanks file
+  # RETURN : array of genbank files
+  def split_genbank outdir, multigbk
+
+    Dir.mkdir("#{outdir}/single-genbank")if ! Dir.exists?("#{outdir}/single-genbank")
+    File.open(multigbk,"r") do |f|
+      fopen = nil
+      while l = f.gets
+        if l[0..9] == "DEFINITION"
+          file_name = l.chomp.split(";")[2].gsub("seqhdr","").delete("\"").delete("=").split(" ")[0]
+          outseq, seq_length = print_sequence_for_gbk @single_fasta[file_name]
+          spacer = " " * (20-seq_length.to_s.length)
+          date = DateTime.now
+          month = Date::ABBR_MONTHNAMES[date.month]
+          day = "%02d" % date.day
+          year = date.year
+          locus = "LOCUS       #{file_name}#{spacer}#{seq_length.to_s} bp    DNA     linear   BCT #{day}-#{month}-#{year}\n"
+          fopen = File.open("#{outdir}/single-genbank/#{file_name}.gbk", "w")
+          fopen.write(locus)
+          fopen.write(l)
+        elsif l[0..1] == "//"
+          fopen.write(outseq)
+          fopen.close
+        else
+          fopen.write(l)
+        end
+      end
+    end
+
+  end
+
+
+  # Utility function to print the sequence to the end of a gbk file
+  def print_sequence_for_gbk seq
+
+    outseq = "ORIGIN\n"
+    # puts "ORIGIN"
+
+    ntNum = 0
+    sequence = seq.seq.downcase
+
+    nt_left = true
+    it = 0
+
+    while nt_left
+
+      if sequence.length > it+60
+        nt_to_add = sequence[it..(it+59)]
+        # printf "%9s ", (ntNum - l.size + 2)
+        outseq += "%9s " % (it+1)
+        outseq += nt_to_add.scan(/.{1,10}/).join(" ")
+        outseq += "\n"
+        it += 60
+      else
+        nt_to_add = sequence[it..sequence.length-1]
+        outseq += "%9s " % (it+1)
+        outseq += nt_to_add.scan(/.{1,10}/).join(" ")
+        outseq += "\n"
+        outseq += "//"
+        nt_left = false
+      end
+
+    end
+    
+    return outseq, sequence.length
+
+  end
+
+
+  # extract protein and gene names
+  def extract_cds_names
+
+    prot_ids = {}
+    flatfile = Bio::FlatFile.auto(@prodigal_files[:proteins])
+    flatfile.each_entry do |entry|
+      prot_id = entry.definition.split(" ")[0]
+      contig = prot_id.split("_")[0..-2].join("_")
+      if !prot_ids.has_key? contig
+        prot_ids[contig] = []
+      end
+      prot_ids[contig] << prot_id
+    end
+
+    prot_ids.each do |k,prot_array|
+      prot_array.sort! { |a,b| a.split("_")[-1].to_i <=> b.split("_")[-1].to_i }
+    end
+
+    @prodigal_files[:prot_ids_by_contig] = prot_ids
+
+  end
+
+
+  private :extract_cds_names # :split_fasta, :split_genbank
+
+end

data/lib/bacterial-annotator/genbank-manip.rb

@@ -0,0 +1,208 @@
+# -*- coding: utf-8 -*-
+# author:  	maxime déraspe
+# email:	maxime@deraspe.net
+# review:  	
+# date:    	15-02-24
+# version: 	0.0.1
+# licence:  	
+
+
+
+class GenbankManip
+
+  attr_accessor :gbk, :coding_seq, :cds_file
+
+  # Initialize then genbank file
+  def initialize gbk_file, outdir
+
+    @gbk_file = gbk_file
+    if ! File.exists? @gbk_file
+      fetch_ncbi_genome(@gbk_file, outdir)
+      @gbk_file = "#{outdir}/#{gbk_file}.gbk"
+      # @gbk_file += ".gbk"
+    end
+
+    flat_gbk = Bio::FlatFile.auto(@gbk_file)
+
+    # Check if gbk is valid
+    if flat_gbk.dbclass != Bio::GenBank    
+      abort "Aborting : The input #{@gbk_file} is not a valid genbank file !"
+    else
+      @gbk = flat_gbk.next_entry
+    end
+
+    @bioseq = @gbk.to_biosequence
+
+  end
+
+
+  # Prepare CDS/proteins
+  def get_cds
+
+    if @coding_seq == nil
+
+      @coding_seq = {}
+
+      # Iterate over each CDS
+      @gbk.each_cds do |ft|
+        ftH = ft.to_hash
+        loc = ft.locations
+        gene = []
+        product = []
+        protId = ""
+        if ftH.has_key? "pseudo"
+          next
+        end
+        gene = ftH["gene"] if !ftH["gene"].nil?
+        product = ftH["product"] if !ftH["product"].nil?
+        protId = ftH["protein_id"][0] if !ftH["protein_id"].nil?
+        locustag = ftH["locus_tag"][0] if !ftH["locus_tag"].nil?
+        dna = get_DNA(ft,@bioseq)
+        if ftH.has_key? "translation"
+          pep = ftH["translation"][0] if !ftH["translation"].nil?
+        else
+          pep = ""
+        end
+
+        pepBioSeq = Bio::Sequence.auto(pep)
+
+        @coding_seq[protId] = {location: loc,
+                               locustag: locustag,
+                               gene: gene[0],
+                               product: product[0],
+                               bioseq: pepBioSeq }
+      end
+
+    end
+
+    @coding_seq
+
+  end
+
+
+  # Print CDS to files
+  # RETURN : cds_file path
+  def write_cds_to_file outdir
+
+    cds_file = "#{@gbk.accession}.pep"
+    if @coding_seq == nil
+      get_cds
+    end
+
+    File.open("#{outdir}/#{cds_file}", "w") do |fwrite|
+      @coding_seq.each_key do |k|
+        seqout = @coding_seq[k][:bioseq].output_fasta("#{k}",60)
+        fwrite.write(seqout)
+      end
+    end
+
+    @cds_file = "#{outdir}/" + cds_file
+
+  end
+
+
+  # add annotation to a genbank file produced by prodigal
+  def add_annotation annotations, outdir, mode
+
+    nb_of_added_ft = 0
+    i = 0
+    contig = @gbk.definition.split(";")[2].
+             gsub("seqhdr","").
+             delete("\"").
+             delete("=").
+             split(" ")[0]
+
+    # iterate through 
+    @gbk.features.each_with_index do |cds, ft_index|
+
+      next if cds.feature != "CDS"
+
+      if mode == 0
+        ftArray = []
+      else
+        ftArray = cds.qualifiers
+      end
+
+      i += 1
+      prot_id = contig+"_"+i.to_s
+      hit = nil
+      hit = annotations[prot_id] if annotations.has_key? prot_id
+
+      if hit != nil
+        locus, gene, product, note = nil
+        locus = hit[:locustag]
+        gene = hit[:gene]
+        product = hit[:product]
+        note = hit[:note]
+
+        if gene != nil
+          qGene = Bio::Feature::Qualifier.new('gene', gene)
+          ftArray.push(qGene)
+        end
+
+        if product != nil
+          qProd = Bio::Feature::Qualifier.new('product', product)
+          ftArray.push(qProd)
+        end
+
+        if locus != nil
+          qNote = Bio::Feature::Qualifier.new('note', "correspond to #{locus} locus")
+          ftArray.push(qNote)
+        end
+
+        if note != nil
+          qNote = Bio::Feature::Qualifier.new('note', note)
+          ftArray.push(qNote)
+        end
+
+
+      end
+      cds.qualifiers = ftArray
+
+    end
+
+    File.open("#{outdir}/#{contig}.gbk", "w") do |f| 
+      f.write(@gbk.to_biosequence.output(:genbank))
+    end
+
+    # Bioruby doesn't support gff at this point
+    # File.open("#{outdir}/#{contig}.gff", "w") do |f| 
+    #   f.write(@gbk.to_biosequence.output(:gff))
+    # end
+
+  end
+
+
+  ###################
+  # Private Methods #
+  ###################
+  
+  # Fct: Get dna sequence
+  def get_DNA (cds, seq)
+    loc = cds.locations
+    sbeg = loc[0].from.to_i
+    send = loc[0].to.to_i
+    fasta = Bio::Sequence::NA.new(seq.subseq(sbeg,send))
+    # position = "#{sbeg}..#{send}"
+    if loc[0].strand == -1
+      fasta.reverse_complement!
+    end
+    dna = Bio::Sequence.auto(fasta)
+    return dna
+  end
+
+
+  # Fetch genbank genome from NCBI
+  def fetch_ncbi_genome refgenome_id, outdir
+    Bio::NCBI.default_email = 'default@default.com'
+    ncbi = Bio::NCBI::REST.new
+    genbankstring = ncbi.efetch(refgenome_id, {"db"=>'nucleotide', "rettype"=>'gb'})
+    File.open("#{outdir}/#{refgenome_id}.gbk", "w") do |f| 
+      f.write(genbankstring)
+    end
+  end
+
+  private :fetch_ncbi_genome, :get_DNA
+
+  
+end                             # end of Class

data/lib/bacterial-annotator/remote-ncbi.rb

@@ -0,0 +1,200 @@
+# -*- coding: utf-8 -*-
+# author:  	maxime déraspe
+# email:	maxime@deraspe.net
+# review:  	
+# date:    	15-02-24
+# version: 	0.0.1
+# licence:  	
+
+require 'mechanize'
+require 'open-uri'
+require 'bio'
+
+class RemoteNCBI
+
+  attr_reader :aln_hits, :db, :xmloutput
+
+  # initialize stuff for a remote ncbi run
+  def initialize db, seq_file, outfile, pidentity
+
+    if ! ["swissprot", "refseq_protein", "nr"].include? db
+      @db = "bad database"
+    else
+      @db = db
+    end
+
+    url = 'http://blast.ncbi.nlm.nih.gov/Blast.cgi'\
+          '?PROGRAM=blastp&BLAST_PROGRAMS=blastp'\
+          '&PAGE_TYPE=BlastSearch&SHOW_DEFAULTS=on'\
+          '&LINK_LOC=blasthome'
+
+    @seq_file = seq_file
+    @outfile = outfile
+    @resultURI = submit_blast url
+    @pidentity = pidentity
+
+    if @resultURI != ""
+      @xmloutput = ""
+      @valid = validate_output
+    else
+      @valid = false
+    end
+
+  end                           # end of method
+
+
+  # submit blast to ncbi
+  def submit_blast ncbiURL
+
+    f = @seq_file.split("/")[-1]
+
+    seq_fasta = File.read(@seq_file)
+
+    a = Mechanize.new { |agent|
+      agent.user_agent_alias = 'Linux Firefox'
+      agent.ignore_bad_chunking = true
+    }
+
+    toBreak = 0
+    requestID = ""
+    try = 1
+
+    while requestID == "" and try < 12
+
+      begin
+
+        a.get(ncbiURL) do |page|
+
+          search = page.form_with(:name => 'searchForm') { |form|
+            form.textareas[0].value = File.read(@seq_file)
+            form.field_with(:name => 'DATABASE').value = @db
+            form.field_with(:name => 'MAX_NUM_SEQ').value = 40
+          }.submit
+
+          search.parser.css('td').each do |td|
+            if toBreak == 1
+              requestID = td.text.gsub(" ","")
+              # puts "breaking because #{requestID}"
+              break
+            end
+            if td.text == "Request ID"
+              toBreak = 1
+            end
+          end
+
+        end
+       
+      rescue
+        try += 1
+        puts "#{try} POST try for #{f}"
+        sleep 3
+      end
+
+    end
+
+    uri_parsed = "http://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Get&RID=#{requestID}"
+    puts "NCBI Blast for #{f}: #{uri_parsed}"
+
+    return URI.parse("http://blast.ncbi.nlm.nih.gov/Blast.cgi?RESULTS_FILE=on&RID=#{requestID}&FORMAT_TYPE=XML&FORMAT_OBJECT=Alignment&CMD=Get")
+
+  end                           # end of method
+
+
+  # validate the xml blast results
+  def validate_output
+
+    xmloutput = ""
+    valid = true
+    finish = false
+
+    while valid and ! finish
+
+      response = Net::HTTP.get_response(@resultURI)
+      body = response.body.split("\n")
+      if body[0] =~ /<?xml version=/
+        xmloutput = body.join("\n")
+        valid = true
+        finish = true
+      else
+        valid = false
+        body.each do |l|
+          if l =~ /Status=/
+            status = l.strip.gsub("Status=", "")
+            if status == "WAITING"
+              valid = true
+            end
+          end
+          break if valid
+        end
+      end
+
+      case @db
+      when 'nr', 'refseq_protein'
+        sleep 30
+      when 'swissprot'
+        sleep 10
+      end
+
+    end
+
+    if finish
+      File.open("#{@outfile}", "w") do |f|
+        f.write(xmloutput)
+      end
+      return finish
+    end
+    valid
+
+  end                           # end of method
+
+  # extract blast results from 
+  def extract_blast_results
+
+    if !@valid
+      @aln_hits = nil
+      return
+    end
+
+    flat = Bio::FlatFile.auto("#{@outfile}")
+    @aln_hits = {}
+
+    flat.each_entry do |report|
+
+      report.iterations.each do |query_it|
+        prot_id = query_it.query_def.split(" ")[0]
+        query_it.hits.each do |hit|
+          if ! @aln_hits.has_key? prot_id
+            p_identity = hit.identity.to_f/hit.target_len.to_f*100
+            if p_identity >= @pidentity
+              # cleaning product definition
+              definition_clean = hit.definition.split(">")[0]
+              product = definition_clean.
+                        gsub("MULTISPECIES: ","").
+                        gsub(/ \[.*\]/,"").
+                        gsub("RecName: Full=","").
+                        split("; AltName")[0].
+                        split("; Flags:")[0].
+                        split(" ; Short=")[0]
+              gi = hit.hit_id.to_s.split("|")[1]
+              organism = ""
+              definition_clean = hit.definition.split(">")[0]
+              if ! definition_clean[/\[.*\]/].nil?                
+                organism = definition_clean[/\[.*\]/].gsub("[","").gsub("]","")
+              end
+              @aln_hits[prot_id] = {
+                pId: hit.identity.to_f/hit.target_len.to_f*100,
+                length: hit.target_len.to_i,
+                evalue: hit.evalue,
+                score: hit.bit_score.to_f,
+                hits: [{gi: gi, product: product, org: organism}]
+              }
+            end
+          end
+        end
+      end
+    end
+
+  end                           # end of method
+
+
+end                             # end of class