rbfam 0.1.3 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 69c97958b2246fffbb4915ee3d60bcd5161c15dd
-  data.tar.gz: 6cbba5157ab211ee188328d1fed2a6b83a07f32b
+  metadata.gz: a266530dd38f3e5f9505a9f4ed0ce6ca4ccdea89
+  data.tar.gz: 2d0376d8ba04b9afe22a7d434f75d9e8bc17a4d2
 SHA512:
-  metadata.gz: 81d8588a7aae75b41d777409740a0db75107ec0b7e5ba059110511da286fc50c1c40314c8471120a862e73e27657146ca37dd8db4f62b70d7d8b02a198526511
-  data.tar.gz: d07932c2d7981fa90f9146554a4c81cdd5497e367c870d13f40bae445bb422affb904a13fef6caaab559fd130d633cf35e5891e9d4a2e2ac8ed412d0122fb7c0
+  metadata.gz: e1534185f816b75dc489fd5b92bb37f5e133a1cf2c6f9c7dbb65fda8184bb8cd7d7ffb5b40e6ca10a608b75355a209246ad41843e9def666f8c9a07422d846a7
+  data.tar.gz: cbdf33c78a5a248a5eb59a02a39affaa6b3ed66fd51651a68e8ac0db18c2ed22157b891a4a2a5c16d37fff48a4c0a8d20a37d008b5e2da251c16639f914402af

data/.gitignore

@@ -0,0 +1,19 @@
+*.ps
+*.gem
+*.rbc
+.bundle
+.config
+coverage
+InstalledFiles
+lib/bundler/man
+pkg
+rdoc
+spec/reports
+test/tmp
+test/version_tmp
+tmp
+
+# YARD artifacts
+.yardoc
+_yardoc
+doc/

data/Gemfile

@@ -0,0 +1,4 @@
+source 'https://rubygems.org'
+
+# Specify your gem's dependencies in wrnap.gemspec
+gemspec

data/Gemfile.lock

@@ -0,0 +1,95 @@
+PATH
+  remote: .
+  specs:
+    rbfam (0.2.0)
+      activerecord
+      activesupport
+      bio (~> 1.4, >= 1.4.2)
+      bio-stockholm (~> 0.0, >= 0.0.1)
+      entrez (~> 0.5, >= 0.5.8.1)
+      httparty (~> 0.8, >= 0.8.3)
+      mysql2 (~> 0.3, >= 0.3.14)
+      nokogiri (~> 1.6, >= 1.6.1)
+      parallel (~> 1.3, >= 1.3.2)
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    actionpack (3.2.21)
+      activemodel (= 3.2.21)
+      activesupport (= 3.2.21)
+      builder (~> 3.0.0)
+      erubis (~> 2.7.0)
+      journey (~> 1.0.4)
+      rack (~> 1.4.5)
+      rack-cache (~> 1.2)
+      rack-test (~> 0.6.1)
+      sprockets (~> 2.2.1)
+    activemodel (3.2.21)
+      activesupport (= 3.2.21)
+      builder (~> 3.0.0)
+    activerecord (3.2.21)
+      activemodel (= 3.2.21)
+      activesupport (= 3.2.21)
+      arel (~> 3.0.2)
+      tzinfo (~> 0.3.29)
+    activesupport (3.2.21)
+      i18n (~> 0.6, >= 0.6.4)
+      multi_json (~> 1.0)
+    arel (3.0.3)
+    bio (1.4.3.0001)
+    bio-stockholm (0.0.1)
+    builder (3.0.4)
+    entrez (0.5.8.1)
+      httparty
+    erubis (2.7.0)
+    hike (1.2.3)
+    httparty (0.13.3)
+      json (~> 1.8)
+      multi_xml (>= 0.5.2)
+    i18n (0.6.11)
+    journey (1.0.4)
+    json (1.8.1)
+    mini_portile (0.6.1)
+    multi_json (1.10.1)
+    multi_xml (0.5.5)
+    mysql2 (0.3.17)
+    nokogiri (1.6.5)
+      mini_portile (~> 0.6.0)
+    parallel (1.3.3)
+    rack (1.4.5)
+    rack-cache (1.2)
+      rack (>= 0.4)
+    rack-ssl (1.3.4)
+      rack
+    rack-test (0.6.2)
+      rack (>= 1.0)
+    railties (3.2.21)
+      actionpack (= 3.2.21)
+      activesupport (= 3.2.21)
+      rack-ssl (~> 1.3.2)
+      rake (>= 0.8.7)
+      rdoc (~> 3.4)
+      thor (>= 0.14.6, < 2.0)
+    rake (10.4.2)
+    rdoc (3.12.2)
+      json (~> 1.4)
+    sprockets (2.2.3)
+      hike (~> 1.2)
+      multi_json (~> 1.0)
+      rack (~> 1.0)
+      tilt (~> 1.1, != 1.3.0)
+    standalone_migrations (2.1.5)
+      activerecord (~> 3.2)
+      railties (~> 3.2)
+      rake (~> 10.0)
+    thor (0.19.1)
+    tilt (1.4.1)
+    tzinfo (0.3.42)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  rbfam!
+  standalone_migrations (~> 2.1, >= 2.1.5)

data/Rakefile

@@ -0,0 +1,3 @@
+require "bundler/gem_tasks"
+require "standalone_migrations"
+StandaloneMigrations::Tasks.load_tasks

data/db/config.yml

@@ -0,0 +1,5 @@
+development:
+  adapter: mysql2
+  database: rbfam
+  username: root
+  password: 

data/db/migrate/20140114232757_add_alignment_table.rb

@@ -0,0 +1,14 @@
+class AddAlignmentTable < ActiveRecord::Migration
+  def up
+    create_table :alignments do |table|
+      table.timestamps
+      table.belongs_to :family
+      table.text :stockholm, limit: 4294967295
+      table.text :consensus_structure
+    end
+  end
+
+  def down
+    drop_table :alignments
+  end
+end

data/db/migrate/20140114232805_add_sequence_table.rb

@@ -0,0 +1,17 @@
+class AddSequenceTable < ActiveRecord::Migration
+  def up
+    create_table :sequences do |table|
+      table.timestamps
+      table.belongs_to :alignment
+      table.string :accession
+      table.text :stripped_sequence
+      table.text :alignment_sequence
+      table.integer :from
+      table.integer :to
+    end
+  end
+
+  def down
+    drop_table :sequences
+  end
+end

data/db/migrate/20140114232810_add_family_table.rb

@@ -0,0 +1,15 @@
+class AddFamilyTable < ActiveRecord::Migration
+  def up
+    create_table :families do |table|
+      table.timestamps
+      table.string :name
+      table.string :description
+    end
+    
+    add_index :families, :name, unique: true
+  end
+
+  def down
+    drop_table :families
+  end
+end

data/db/schema.rb

@@ -0,0 +1,44 @@
+# encoding: UTF-8
+# This file is auto-generated from the current state of the database. Instead
+# of editing this file, please use the migrations feature of Active Record to
+# incrementally modify your database, and then regenerate this schema definition.
+#
+# Note that this schema.rb definition is the authoritative source for your
+# database schema. If you need to create the application database on another
+# system, you should be using db:schema:load, not running all the migrations
+# from scratch. The latter is a flawed and unsustainable approach (the more migrations
+# you'll amass, the slower it'll run and the greater likelihood for issues).
+#
+# It's strongly recommended to check this file into your version control system.
+
+ActiveRecord::Schema.define(:version => 20140114232810) do
+
+  create_table "alignments", :force => true do |t|
+    t.datetime "created_at",                                :null => false
+    t.datetime "updated_at",                                :null => false
+    t.integer  "family_id"
+    t.text     "stockholm",           :limit => 2147483647
+    t.text     "consensus_structure"
+  end
+
+  create_table "families", :force => true do |t|
+    t.datetime "created_at",  :null => false
+    t.datetime "updated_at",  :null => false
+    t.string   "name"
+    t.string   "description"
+  end
+
+  add_index "families", ["name"], :name => "index_families_on_name", :unique => true
+
+  create_table "sequences", :force => true do |t|
+    t.datetime "created_at",         :null => false
+    t.datetime "updated_at",         :null => false
+    t.integer  "alignment_id"
+    t.string   "accession"
+    t.text     "stripped_sequence"
+    t.text     "alignment_sequence"
+    t.integer  "from"
+    t.integer  "to"
+  end
+
+end

data/lib/rbfam.rb

@@ -1,15 +1,22 @@
-require "json"
+require "active_record"
+require "active_support/inflector"
 require "bio"
-require "vienna_rna"
+require "bio-stockholm"
 require "entrez"
 require "httparty"
-require "active_support/inflector"
+require "json"
+require "mysql2"
+require "parallel"
 
 %W|helpers modules|.each do |folder|
   Dir[File.join(File.dirname(__FILE__), "rbfam", folder, "*.rb")].each { |name| require "rbfam/#{folder}/#{File.basename(name, '.rb')}" }
 end
 
 module Rbfam
+  def self.connect(config: nil)
+    ActiveRecord::Base.establish_connection(config || YAML.load_file(File.join(File.dirname(__FILE__), "..", "db", "config.yml"))["development"])
+  end
+  
   def self.script(name)
     require "rbfam/scripts/#{File.basename(name, '.rb')}"
   end

data/lib/rbfam/modules/alignment.rb

@@ -1,46 +1,44 @@
 module Rbfam
   class Alignment
     include Rbfam::CommonHelpers
-    
+
     LINE_REGEXP = /^([\w\.]+)\/(\d+)\-(\d+)\s+([AUGC\.]+)$/
-    
+
     attr_reader :family, :seed
-    
+
     def initialize(family)
       @family = family
     end
-    
+
     def entries(options = {})
       options = { alignment: :seed, limit: false }.merge(options)
-      
-      @parsed_entries ||= (pull_from_server(options[:alignment]) || "").split(/\n/).reject do |line| 
+
+      @parsed_entries ||= (pull_from_server(options[:alignment]) || "").split(/\n/).reject do |line|
         line =~ /^#/
-      end.select do |line| 
+      end.select do |line|
         line =~ LINE_REGEXP
       end[options[:limit] ? 0...options[:limit] : 0..-1].map(&method(:parse_line)).tap do
         @seed = options[:alignment] == :seed
       end
     end
-    
+
     def save_entries!
       entries.each { |sequence| sequence.save!(seed: seed) }
     end
-    
+
     def load_entries!(options = {})
-      options = { extended: false }.merge(options)
-      
       @parsed_entries = family.load_entries!(options)
     end
-    
+
     private
-    
+
     def pull_from_server(alignment)
       url = "http://rfam.sanger.ac.uk/family/alignment/download/format?acc=%s&alnType=%s&nseLabels=1&format=pfam&download=0" % [
         family.family_name,
         alignment
       ]
       puts "GET: %s" % url unless @reponse
-      
+
       @response ||= if (party = HTTParty.get(url)).response.code == "200"
         puts "RESPONSE: 200 OK"
         party.parsed_response
@@ -52,11 +50,11 @@ module Rbfam
         ])
       end
     end
-    
+
     def parse_line(line)
       line_match = line.match(LINE_REGEXP)
-      
+
       Rbfam::Sequence.new(family, line_match[1], line_match[2].to_i, line_match[3].to_i, autoload: true)
     end
   end
-end
+end

data/lib/rbfam/modules/family.rb

@@ -1,56 +1,54 @@
 module Rbfam
   class Family
     include Rbfam::CommonHelpers
-    
+
     attr_reader :family_name, :description
-    
+
     class << self
       def method_missing(name, *args, &block)
         READABLE[name.to_s] ? new(READABLE[name.to_s]) : super
       end
     end
-    
+
     def initialize(family_name, description = "")
       @family_name = family_name
       @description = description
     end
-    
+
     def id
       ("%s %s" % [family_name, description.downcase]).strip.gsub(/\W+/, "_")
     end
-    
+
     def alignment
       Rbfam::Alignment.new(self)
     end
-    
+
     def entries(options = {})
       options = { limit: false }.merge(options)
-      
+
       @parsed_entries ||= pull_from_server.split(/\n/).reject { |line| line =~ /^#/ }[options[:limit] ? 0...options[:limit] : 0..-1].map(&method(:parse_line))
     end
-    
+
     def load_entries!(options = {})
-      options = { extended: false, seed: true }.merge(options)
-      
       Rbfam.script("sequences_in_mysql")
-      
-      @parsed_entries = SequenceTable.where({ family: family_name }.merge(options)).map do |entry|
+
+      Rbfam::DB::Sequence.joins(alignment: :family).where(families: { name: family_name }.merge(options)).map do |entry|
         entry.to_rbfam(self)
       end
     end
-    
+
     def save_entries!
       entries.each(&:save!)
     end
-    
+
     private
-    
+
     def pull_from_server
       # It isn't the greatest design pattern to memoize a block where a branch has unmanaged exception raising, but for my uses that should never
       # happen and needs to blow up hard if it does.
       url = "http://rfam.sanger.ac.uk/family/regions?entry=%s" % family_name
       puts "GET: %s" % url unless @reponse
-      
+
       @reponse ||= if (party = HTTParty.get(url)).response.code == "200"
         puts "RESPONSE: 200 OK"
         party.parsed_response
@@ -62,11 +60,11 @@ module Rbfam
         ])
       end
     end
-    
+
     def parse_line(line)
       split_line = line.split(/\t/)
-      
+
       Rbfam::Sequence.new(self, split_line[0], split_line[2].to_i, split_line[3].to_i, autoload: true)
     end
   end
-end
+end

data/lib/rbfam/modules/sequence.rb

@@ -1,5 +1,5 @@
 module Rbfam
-  class Sequence
+  class Sequence < ActiveRecord::Base
     attr_reader :family, :accession, :from, :to, :coord_options
 
     def initialize(family, accession, from, to, options = {})
@@ -66,25 +66,15 @@ module Rbfam
       !plus_strand?
     end
 
-    def sequence
+    alias_method :seq, def sequence
       @raw_sequence ||= Rbfam::Utils.rna_sequence_from_entrez(accession, up_coord, coord_window)
       @raw_sequence   = minus_strand? ? @raw_sequence.complement : @raw_sequence
     end
     
-    alias :seq :sequence
-    
-    def mfe_structure
-      @mfe_structure ||= ViennaRna::Fold.run(seq).structure
-    end
-    
     def description
       ("%s %s %s" % [accession, from, to]).gsub(/\W+/, "_")
     end
     
-    def fftbor
-      @fftbor ||= ViennaRna::Fftbor.run(seq: seq, str: mfe_structure)
-    end
-    
     def extend!(coord_options = {})
       tap do
         @extended      = true
@@ -109,7 +99,7 @@ module Rbfam
           case [coord_options[:direction], strand]
           when [3, :plus], [5, :minus] then Range.new(range.min, range.max + coord_options[:length])
           when [5, :plus], [3, :minus] then Range.new(range.min - coord_options[:length], range.max)
-          else puts "WARNING: value for :direction key in sequence retreival needs to be one of 5, 3, :both - found (%s)" % coord_options[:direction].inspect
+          else puts "WARNING: value for :direction key in sequence retrieval needs to be one of 5, 3, :both - found (%s)" % coord_options[:direction].inspect
           end
         end
       else
@@ -121,4 +111,4 @@ module Rbfam
       "#<Rbfam::Sequence #{description} #{seq[0, 20] + ('...' if seq.length > 20)}>"
     end
   end
-end
+end