rbbt-phgx 0.1.0 → 0.2.0

data/lib/phgx.rb

@@ -1,16 +1,7 @@
 require 'rbbt-util'
-require 'rbbt/util/pkg_config'
-require 'rbbt/util/pkg_data'
-require 'rbbt/util/pkg_software'
-require 'rbbt/util/open'
-require 'rbbt/util/tmpfile'
-require 'rbbt/util/filecache'
+require 'rbbt/util/resource'
 
 module PhGx
-  extend PKGConfig
-  extend PKGData
-  extend PKGSoftware
-
-  self.load_cfg(%w(datadir), "datadir: #{File.join(ENV['HOME'], 'phgx', 'data')}\n")
+  extend Resource
 end
 

data/lib/rbbt/mutation/fireDB.rb

@@ -0,0 +1,11 @@
+require 'phgx'
+require 'rbbt/util/cmd'
+
+module FireDB
+  PhGx.add_software "FireDB" => ['', :directory]
+
+  def self.predict(accession, sequence, mutation)
+    CMD.cmd("perl " + File.join(PhGx.find_software("FireDB"), "firePredText.pl") + " " + [accession, accession, sequence, 10] * " ").read
+  end
+
+end

data/lib/rbbt/mutation/polyphen.rb

@@ -0,0 +1,172 @@
+require 'rbbt/util/open'
+require 'rbbt/util/tsv'
+require 'nokogiri'
+require 'digest/md5'
+
+module Polyphen2
+
+  URL="http://genetics.bwh.harvard.edu/cgi-bin/ggi/ggi2.cgi"
+  URL_BASE="http://genetics.bwh.harvard.edu/"
+
+  module Batch
+    OPTIONS = {
+      "_ggi_project"                      => "PPHWeb2",
+      "_ggi_origin"                       => "query",
+      "_ggi_batch_file"                   => "",
+      "description"                       => "",
+      "NOTIFYME"                          => "",
+      "uploaded_sequences_1"              => "",
+      "description_of_uploaded_sequences" => "",
+      "MODELNAME"                         => "HumDiv",
+      "UCSCDB"                            => "hg19",
+      "SNPFILTER"                         => "1",
+      "SNPFUNC"                           => "m",
+      "_ggi_target_pipeline"              => "Submit Batch",
+    }
+
+    REFRESH_OPTIONS = {
+      "sid"                => "",
+      "_ggi_project"       => "PPHWeb2",
+      "_ggi_origin"        => "manage",
+      "_ggi_target_manage" => "Refresh",
+    }
+
+  def self.predict(query)
+    options = OPTIONS.merge "_ggi_batch" => query
+
+    desc =  Digest::MD5.hexdigest(options.inspect)
+    options["description"] = desc
+
+    ddd desc
+
+    doc = Nokogiri::HTML(Open.read(Polyphen2::URL, :wget_options => {"--post-data" => "'#{options.collect{|k,v| [k,v] * "="} * "&"}'"}, :nocache => true))
+
+    sid = doc.css('input[name=sid]').attr('value')
+
+    options = REFRESH_OPTIONS.merge "sid" => sid
+    finished = false
+
+    view_link = nil
+    while not finished do
+      doc = Nokogiri::HTML(Open.read(Polyphen2::URL, :wget_options => {"--post-data" => "'#{options.collect{|k,v| [k,v] * "="} * "&"}'"}, :nocache => true))
+
+      result_table =  doc.css('body > table')[1].css('table')[2]
+
+      rows = result_table.css('tr')
+
+      row = rows.select{|row| row.css('td').length == 6}.select{|row| row.css('td').last.content.strip == desc}.first
+
+      cells = row.css('td')
+      if cells[2].content =~ /Error/
+        view_link = nil
+        break
+      end
+
+      if cells[1].content =~ /Short/
+        view_link =  cells[1].css('a').attr('href')
+        break
+      end
+
+      sleep 1
+    end
+
+    return nil if view_link.nil?
+
+    tsv = TSV.new Open.open(Polyphen2::URL_BASE + view_link, :nocache => true), :double, :key => 'acc', :merge => true
+
+    return tsv
+  end
+
+
+  end
+
+
+  OPTIONS = {
+    "ContAllHits"        => 0,
+    "ContThresh"         => 6,
+    "Map2Mismatch"       => 0,
+    "MaxHitGaps"         => 20,
+    "MinHitIde"          => 0.5,
+    "MinHitLen"          => 100,
+    "SortByIde"          => 1,
+    "StructAllHits"      => 0,
+    "_ggi_jpover"        => 1,
+    "_ggi_origin"        => "query",
+    "_ggi_project"       => "PPHWeb2",
+    "_ggi_target_submit" => "submit",
+    "accid"              => "A6NFZ4",
+    "description"        => "",
+    "seqpos"             => "34",
+    "seqres"             => "",
+    "seqvar1"            => "Y",
+    "seqvar2"            => "D",
+    "Submit"             => "Submit+Query",
+  }
+
+  REFRESH_OPTIONS = {
+    "sid"                => "",
+    "_ggi_project"       => "PPHWeb2",
+    "_ggi_origin"        => "manage",
+    "_ggi_target_manage" => "Refresh",
+  }
+
+  def self.parse_mutation(mutation)
+    mutation.match(/([A-Z])(\d+)([A-Z])/i).values_at 1,2,3
+  end
+
+  def self.predict(accession, mutation)
+    reference, pos, substitution = parse_mutation(mutation)
+
+    options = OPTIONS.merge "accid" => accession, "seqpos" => pos, "seqvar1" => reference, "seqvar2" => substitution
+
+    desc =  Digest::MD5.hexdigest(options.inspect)
+    options["description"] = desc
+
+    doc = Nokogiri::HTML(Open.read(URL, :wget_options => {"--post-data" => "'#{options.collect{|k,v| [k,v] * "="} * "&"}'"}, :nocache => true))
+
+    sid = doc.css('input[name=sid]').attr('value')
+
+    options = REFRESH_OPTIONS.merge "sid" => sid
+    finished = false
+
+    view_link = nil
+    while not finished do
+      doc = Nokogiri::HTML(Open.read(URL, :wget_options => {"--post-data" => "'#{options.collect{|k,v| [k,v] * "="} * "&"}'"}, :nocache => true))
+
+      result_table =  doc.css('body > table')[1].css('table')[2]
+
+      rows = result_table.css('tr')
+
+      row = rows.select{|row| row.css('td').length == 6}.select{|row| row.css('td').last.content.strip == desc}.first
+
+      cells = row.css('td')
+      if cells[2].content =~ /Error/
+        view_link = nil
+        break
+      end
+
+      if cells[1].content =~ /View/
+        view_link =  cells[1].css('a').attr('href')
+        break
+      end
+
+      sleep 1
+    end
+
+    return nil if view_link.nil?
+
+
+    doc = Nokogiri::HTML(Open.read(URL_BASE + view_link, :nocache => true))
+
+    para = doc.css('div#HumDivConf > p').first
+    div_prediction = para.css('span').first.content
+    div_score      = para.css('b').first.content
+
+    para = doc.css('div#HumVarConf > p').first
+    var_prediction = para.css('span').first.content
+    var_score      = para.css('b').first.content
+
+    return [div_prediction, div_score, var_prediction, var_score]
+
+  end
+end

data/lib/rbbt/mutation/sift.rb

@@ -0,0 +1,112 @@
+require 'rbbt/util/open'
+require 'nokogiri'
+module SIFT
+  URL_AMINOACID="http://sift.jcvi.org/sift-bin/SIFT_pid_subst_all_submit.pl"
+  URL_GENOMIC="http://sift.jcvi.org/sift-bin/SIFT_feed_to_chr_coords.pl"
+
+  def self.predict_aminoacid_mutation(accession, mutations)
+    doc = Nokogiri::HTML(Open.read(URL_AMINOACID, :wget_options => {"--post-data" => "'GI=#{[accession, mutations].flatten * ","}&sequences_to_select=BEST&seq_identity_filter=90'"}, :nocache => false))
+
+    rows = []
+    doc.css('tr').each do |row|
+      rows << row.css('td').collect{|cell| cell.content}
+    end
+
+    rows.shift
+
+    if Array === mutations
+      rows
+    else
+      rows.first
+    end
+  end
+
+  def self.predict_aminoacid_mutation_batch(mutations)
+    data = case
+             when String === mutations
+               mutations
+             when Array === mutations
+               mutations.collect{|p| p * ", "} * "\n" if Array === mutations
+             end
+
+    doc = Nokogiri::HTML(Open.read(URL_AMINOACID, :wget_options => {"--post-data" => "'GI=#{data}&sequences_to_select=BEST&seq_identity_filter=90'"}, :nocache => false))
+
+    rows = []
+    doc.css('tr').each do |row|
+      rows << row.css('td').collect{|cell| cell.content}
+    end
+
+    rows.shift
+
+    if Array === mutations 
+      rows
+    else
+      rows.first
+    end
+  end
+
+  def self.parse_genomic_mutation(mutation)
+    mutation.match(/(\d+):(\d+):(1|-1):([A-Z])\/([A-Z])/).values_at 1,2,3,4,5
+  end
+
+  def self.add_predictions(tsv)
+    raise "Input not TSV" unless TSV === tsv
+
+    raise "Field 'Refseq Protein ID' Not in TSV" unless tsv.fields.include? "Refseq Protein ID" 
+
+    raise "Field 'Protein Mutation' Not in TSV" unless tsv.fields.include? "Protein Mutation" 
+
+    data = []
+    tsv.through :key, ["Refseq Protein ID", "Protein Mutation"] do |key,values|
+      refseqs, mutations = values
+      mutations = mutations.reject{|mutation| mutation[0] == mutation[-1]}
+      next if refseqs.nil? or refseqs.compact.reject{|v| v.nil? or v.empty?}.empty? or mutations.empty?
+
+      refseqs.compact.uniq.each do |refseq|
+        data << [refseq, mutations]
+      end
+    end
+
+    data.sort!
+
+
+    predictions = {}
+    predict_aminoacid_mutation_batch(data).each{|values| predictions[values[0] + ":" << values[1]] = values.values_at 3,4,5,6}
+
+    refseq_field = tsv.identify_field "Refseq Protein ID" 
+    protein_field = tsv.identify_field "Protein Mutation" 
+ 
+    tsv.add_field "SIFT:Prediction" do |key,values|
+      refseqs = if refseq_field === :key
+                  [key]
+                else
+                  values[refseq_field] || []
+                end
+
+      next if refseqs.compact.reject{|v| v.nil? or v.empty?}.empty?
+
+      mutations = values[protein_field]
+
+      refseqs.zip(mutations).collect do |refseq,mutation|
+        case
+        when (mutation.nil? or mutation.empty?)
+          "No Prediction"
+        when mutation[0] == mutation[-1]
+          "TOLERATED"
+        when (refseq.nil? or refseq.empty?)
+          "No Prediction"
+        else
+          list = predictions[refseq + ":" << mutation]
+          if list.nil?
+            "No Prediction"
+          else
+            list.first
+          end
+        end
+      end
+    end
+
+    tsv
+  end
+
+end

data/lib/rbbt/mutation/snps_and_go.rb

@@ -0,0 +1,61 @@
+require 'rbbt/util/log'
+require 'rbbt/util/open'
+module SNPSandGO
+  URL="http://snps-and-go.biocomp.unibo.it/cgi-bin/snps-and-go/runpred.cgi?uniprot=#ACCESSION#&position=#POSITION#&wild-type=#REFERENCE#&substituting=#SUBSTITUTION#"
+
+  def self.parse_mutation(mutation)
+    mutation.match(/([A-Z])(\d+)([A-Z])/i).values_at 1,2,3
+  end
+
+  def self.predict(accession, mutation)
+    reference, pos, substitution = parse_mutation(mutation)
+
+    url = URL.sub(/#ACCESSION#/,accession).sub(/#POSITION#/, pos).sub(/#REFERENCE#/,reference).sub(/#SUBSTITUTION#/,substitution)
+
+    res = Open.read(url)
+
+    raise "Error in prediction" unless res =~ /RESULTS/
+
+    res.match(/Position\s+WT\s+NEW\s+Effect\s+RI\n\s+\d+\s+[A-Z]\s+[A-Z]\s+(\w+)\s+(\d+)/).values_at 1,2
+  end
+
+  def self.add_predictions(tsv)
+    raise "Input not TSV" unless TSV === tsv
+
+    uniprot_field = tsv.identify_field "UniProt/SwissProt Accession" 
+    raise "Field 'UniProt/SwissProt Accession' Not in TSV" if uniprot_field.nil?
+
+    protein_field = tsv.identify_field "Protein Mutation" 
+    raise "Field 'Protein Mutation' Not in TSV" if protein_field.nil?
+
+
+    tsv.add_field "SNPs&GO:Prediction" do |key,values|
+      uniprots = if uniprot_field === :key
+                   [key]
+                 else
+                   values[uniprot_field] || []
+                 end
+
+      mutations = values[protein_field]
+
+      uniprots.zip(mutations).collect{|uniprot,mutation| 
+        case
+        when mutation.nil?
+          "No Prediction" 
+        when mutation[0] == mutation[-1]
+            "Neutral"
+        when (uniprot.nil? or uniprot.empty?)
+          "No Prediction" 
+        else
+          begin
+            SNPSandGO.predict(uniprot, mutation).first
+          rescue
+            "No Prediction"
+          end
+        end
+      }
+    end
+
+    tsv
+  end
+end

data/lib/rbbt/sources/biogrid.rb

@@ -0,0 +1,11 @@
+require 'phgx'
+
+module Biogrid
+  extend Resource
+  data_module PhGx
+
+  ["Hsa", "Rno", "Sce"].each do |organism|
+    module_eval "#{ organism } = with_key '#{organism}'"
+  end
+
+end

data/lib/rbbt/sources/cancer.rb

@@ -1,12 +1,7 @@
 require 'phgx'
-require 'rbbt/util/data_module'
 
 module Cancer
-  PhGx.add_datafiles :anais_annotations => ['Cancer', 'Cancer/anais-annotations.txt'],
-    :anais_interactions => ['Cancer', 'Cancer/anais-interactions.txt']
-
-  PKG = PhGx
-  extend DataModule
+  extend Resource
+  relative_to Rbbt, "share/Cancer"
+  @namespace = "Cancer"
 end
-
-if __FILE__ == $0 then NCI.all end

data/lib/rbbt/sources/dbsnp.rb

@@ -0,0 +1,6 @@
+require 'phgx'
+
+module DBSNP
+  extend Resource
+  data_module PhGx
+end

data/lib/rbbt/sources/hprd.rb

@@ -0,0 +1,6 @@
+require 'phgx'
+
+module HPRD
+  extend Resource
+  data_module PhGx
+end

data/lib/rbbt/sources/kegg.rb

@@ -1,9 +1,6 @@
-require 'rbbt/util/data_module'
 require 'phgx'
 
 module KEGG
-  PKG = PhGx
-  extend DataModule
+  extend Resource
+  data_module(PhGx)
 end
-
-if __FILE__ == $0 then  KEGG.all end

data/lib/rbbt/sources/matador.rb

@@ -1,9 +1,6 @@
-require 'rbbt/util/data_module'
 require 'phgx'
 
 module Matador
-  PKG = PhGx
-  extend DataModule
+  extend Resource
+  data_module PhGx
 end
-
-if __FILE__ == $0 then  Matador.all end