bacterial-annotator 0.7.0 → 0.7.1

data/lib/bacterial-identificator.rb

@@ -10,6 +10,9 @@ require 'bio'
 require 'fileutils'
 require 'parallel'
 require 'helper'
+require 'json'
+require 'pp'
+
 
 class BacterialIdentificator
 
@@ -20,15 +23,27 @@ class BacterialIdentificator
   def initialize options, root
 
     @root = root
-    @db_path = options[:database]
-    @genomes_list = options[:genomes_list]
+    @mash_file = options[:mash_file]
+    @genome_list = options[:genome_list]
     @proc = options[:proc].to_i
-    p @genomes_list
+    @output=options[:output]
+
+  end
+
+
+  def run_identification
+
+    @genome_hits = {}
+    @genome_list.each do |g|
+      @genome_hits[g] = []
+    end
 
-    @genomes_list.each do |g|
-      mash_genome g
+    Parallel.map(@genome_list, in_threads: @proc) do |g|
+      @genome_hits[g] = mash_genome g
     end
 
+    print_output
+
   end
 
 
@@ -37,28 +52,86 @@ class BacterialIdentificator
     # Reference-ID, Query-ID, Mash-distance, P-value, and Matching-hashes
     # fields = ["hit","query","distance","pvalue","match"]
 
-    results_raw = `#{@root}/mash.linux dist #{@db_path}/species-sequences.msh #{genome}`
+    results_raw = `#{@root}/mash.linux dist #{@mash_file} #{genome}`
     results = []
 
     results_raw.split("\n").each do |l|
       lA = l.chomp.split("\t")
       next if lA[-1].split("/")[0] == '0' # no match
-      results << lA
+      results << (lA[0..0] + lA[2..-1])
+    end
+
+    results_sorted = results.sort {|a,b| a[1] <=> b[1]}
+
+    return results_sorted
+
+  end
+
+  # consensus species model
+  def consensus_reference
+
+    all_hits = {}
+    @genome_hits.each do |g, hits|
+      hits.each do |h|
+        score = h[3].split("/")[0].to_i
+        if ! all_hits.has_key? h[0]
+          all_hits[h[0]] = score
+        else
+          all_hits[h[0]] += score
+        end
+      end
     end
+    return all_hits.sort_by { |k,v| v }.to_h
+
+  end
 
-    results_sorted = results.sort {|a,b| a[2] <=> b[2]}
+  # print json
+  def print_output
 
-    File.open("#{genome}.msh_dist", "w") do |fout|
-      results_sorted.each do |f|
-        fout.write(f.join("\t"))
-        fout.write("\n")
+    case @output.downcase
+    when "csv"
+      @genome_hits.each do |g, hits|
+        hits.each do |h|
+          puts "#{g},#{h.join(',')}"
+        end
+      end
+    when "json"
+      new_genome_hits = {}
+      @genome_hits.each do |g, hits|
+        new_genome_hits[g] = []
+        hits.each do |h|
+          new_genome_hits[g].push(Hash[["hit","distance","e-value","score"].zip(h)])
+        end
+      end
+      puts JSON.pretty_generate({genomes: new_genome_hits, summary: summary})
+    else
+      @genome_hits.each do |g, hits|
+        hits.each do |h|
+          out = h.join("\t")
+          puts "#{g}\t#{out}"
+        end
       end
     end
 
   end
 
+  def summary
 
+    genome_hit_association =  {}
 
+    @genome_hits.each do |g, hits|
+      genome_hit_association[hits[0][0]] = 0 if ! genome_hit_association.has_key? hits[0][0]
+      genome_hit_association[hits[0][0]] += 1
+    end
+
+    population = {
+      consensus: consensus_reference.first[0],
+      genome_hits: genome_hit_association
+    }
+
+    return population
+
+  end
 
-end
 
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: bacterial-annotator
 version: !ruby/object:Gem::Version
-  version: 0.7.0
+  version: 0.7.1
 platform: ruby
 authors:
 - Maxime Deraspe
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-11-06 00:00:00.000000000 Z
+date: 2018-04-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bio
@@ -124,7 +124,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.5.1
+rubygems_version: 2.6.14
 signing_key: 
 specification_version: 4
 summary: Bacterial Annotator