adjustments for latest lazar version

12 files changed, 134 insertions, 23 deletions

diff --git a/scripts/crossvalidation-summary.rb b/scripts/crossvalidation-summary.rb
new file mode 100755
index 0000000..13b0dfa
--- /dev/null
+++ b/scripts/crossvalidation-summary.rb
@@ -0,0 +1,9 @@
+#!/usr/bin/env ruby
+require_relative '../../lazar/lib/lazar'
+include OpenTox
+
+summary = [] 
+model = Model::Validation.find(File.read(ARGV[0]).chomp).crossvalidations.each do |cv|
+  summary << cv.statistics
+end
+puts JSON.pretty_generate(summary)
diff --git a/scripts/drop-database.rb b/scripts/drop-database.rb
new file mode 100755
index 0000000..e263312
--- /dev/null
+++ b/scripts/drop-database.rb
@@ -0,0 +1,5 @@
+#!/usr/bin/env ruby
+require_relative '../../lazar/lib/lazar'
+$mongo.database.drop
+$gridfs = $mongo.database.fs # recreate GridFS indexes
+
diff --git a/scripts/efsa2csv.rb b/scripts/efsa2csv.rb
new file mode 100755
index 0000000..5251cae
--- /dev/null
+++ b/scripts/efsa2csv.rb
@@ -0,0 +1,29 @@
+#!/usr/bin/env ruby
+require 'csv'
+require_relative '../../lazar/lib/lazar.rb'
+
+i = 0
+db = {}
+CSV.foreach(ARGV[0], :encoding => "UTF-8", :col_sep => "\t", :liberal_parsing => true) do |row|
+  #STDERR.puts i if i%100 == 0
+  if i > 0 and row[11] and !row[11].empty? and row[24].match(/Salmonella/i) and ( row[25].match("TA 98") or row[25].match("TA 100") ) and row[33]
+    begin
+      c = OpenTox::Compound.from_smiles(row[11].gsub('"','')).smiles
+    rescue
+      c = OpenTox::Compound.from_inchi(row[12]).smiles # some smiles (row[11]) contain non-parseable characters
+    end
+    db[c] ||= {}
+    db[c][:id] ||= row[2]
+    if row[33].match(/Positiv/i)
+      db[c][:value] = 1 # at least one positive result in TA 98 or TA 100
+    elsif row[33].match(/Negativ/i)
+      db[c][:value] ||= 0
+    end
+  end
+  i += 1
+end
+
+puts "ID,SMILES,Mutagenicity"
+db.each do |s,v|
+  puts [v[:id],s,v[:value]].join ","
+end
diff --git a/scripts/export.rb b/scripts/export.rb
new file mode 100755
index 0000000..afcb51d
--- /dev/null
+++ b/scripts/export.rb
@@ -0,0 +1,7 @@
+#!/usr/bin/env ruby
+require_relative '../../lazar/lib/lazar'
+include OpenTox
+
+ext = File.extname(ARGV[0]).sub('.','')
+dataset = Dataset.find File.read(ARGV[0]).chomp
+dataset.send("to_#{ext}")
diff --git a/scripts/hansen2csv.rb b/scripts/hansen2csv.rb
new file mode 100755
index 0000000..3ae22ee
--- /dev/null
+++ b/scripts/hansen2csv.rb
@@ -0,0 +1,10 @@
+#!/usr/bin/env ruby
+require 'csv'
+require_relative '../../lazar/lib/lazar.rb'
+
+puts "ID,SMILES,Mutagenicity"
+i = 0
+CSV.foreach(ARGV[0]) do |row|
+  puts [row[0],row[5],row[2]].join "," if i > 0
+  i += 1
+end
diff --git a/scripts/import-pubchem.rb b/scripts/import-pubchem.rb
new file mode 100755
index 0000000..c18bc81
--- /dev/null
+++ b/scripts/import-pubchem.rb
@@ -0,0 +1,8 @@
+#!/usr/bin/env ruby
+require_relative '../../lazar/lib/lazar'
+include OpenTox
+
+dataset = Dataset.send("from_pubchem_aid", ARGV[0])
+puts dataset.id.to_s
+
+
diff --git a/scripts/import.rb b/scripts/import.rb
new file mode 100755
index 0000000..f166265
--- /dev/null
+++ b/scripts/import.rb
@@ -0,0 +1,7 @@
+#!/usr/bin/env ruby
+require_relative '../../lazar/lib/lazar'
+include OpenTox
+
+ext = File.extname(ARGV[0]).sub('.','')
+dataset = Dataset.send("from_#{ext}_file", ARGV[0])
+puts dataset.id.to_s
diff --git a/scripts/json2csv.rb b/scripts/json2csv.rb
new file mode 100755
index 0000000..7c61c5a
--- /dev/null
+++ b/scripts/json2csv.rb
@@ -0,0 +1,28 @@
+#!/usr/bin/env ruby
+require_relative '../../lazar/lib/lazar'
+include OpenTox
+
+results = JSON.parse File.read(ARGV[0])
+puts "SMILES,#{File.basename(ARGV[0]).sub("pa_","").capitalize},Probability(0),Probability(1),Nr Neighbors,Warnings"
+results.each do |s,r|
+	if r["value"]
+		puts [
+			s,
+			r["value"],
+			r["probabilities"]["0"],
+			r["probabilities"]["1"],
+			r["neighbors"].size,
+			r["warnings"],
+		].join(",")
+	else
+		r["neighbors"] ? n = r["neighbors"].size : n = nil
+		puts [
+			s,
+			r["value"],
+			nil,
+			nil,
+			n,
+			r["warnings"],
+		].join(",")
+	end
+end
diff --git a/scripts/merge.rb b/scripts/merge.rb
new file mode 100755
index 0000000..5e9dac4
--- /dev/null
+++ b/scripts/merge.rb
@@ -0,0 +1,11 @@
+#!/usr/bin/env ruby
+require_relative '../../lazar/lib/lazar.rb'
+include OpenTox
+
+hansen = Dataset.find File.read(ARGV[0]).chomp
+efsa = Dataset.find File.read(ARGV[1]).chomp
+kazius = Dataset.find File.read(ARGV[2]).chomp
+datasets = [hansen,efsa,kazius]
+map = {"mutagen" => "1", "nonmutagen" => "0"}
+dataset = Dataset.merge datasets: datasets, features: datasets.collect{|d| d.bioactivity_features.first}, value_maps: [nil,nil,map], keep_original_features: true, remove_duplicates: true
+puts dataset.id
diff --git a/scripts/model.rb b/scripts/model.rb
new file mode 100755
index 0000000..b86d33d
--- /dev/null
+++ b/scripts/model.rb
@@ -0,0 +1,7 @@
+#!/usr/bin/env ruby
+require_relative '../../lazar/lib/lazar'
+include OpenTox
+
+dataset = Dataset.find File.read(ARGV[0]).chomp
+model_validation = Model::Validation.from_dataset training_dataset: dataset, prediction_feature: dataset.merged_features.first, species: "Salmonella typhimurium", endpoint: "Mutagenicity", repeats: 3
+puts model_validation.id.to_s
diff --git a/scripts/predict.rb b/scripts/predict.rb
index 9161f45..57d5a8b 100755
--- a/scripts/predict.rb
+++ b/scripts/predict.rb
@@ -2,23 +2,7 @@
 require_relative '../../lazar/lib/lazar'
 include OpenTox
 
-train = Dataset.from_csv_file ARGV[0]
-
-model = Model::LazarClassification.create(training_dataset: train)
-
-n = 0
-results = {}
-File.open(ARGV[1]).each_line do |l|
-  unless n == 0
-    id,smi = l.chomp.split ","
-    c = Compound.from_smiles smi
-    result = model.predict c
-    results[id] = result
-  end
-  n += 1
-end
-
-puts results.to_json
-  #puts result.inspect
-  #rcv = Validation::RepeatedCrossValidation.create(model)
-#puts "#{rcv.id}"
+dataset = Dataset.find(File.read(ARGV[0]).chomp)
+model = Model::Validation.find(File.read(ARGV[1]).chomp)
+predictions = model.predict dataset.compounds # avoid dataset prediction to keep neighbors
+puts predictions.to_json
diff --git a/scripts/repeated_crossvalidation_summary.rb b/scripts/repeated_crossvalidation_summary.rb
index add66a2..067fea8 100755
--- a/scripts/repeated_crossvalidation_summary.rb
+++ b/scripts/repeated_crossvalidation_summary.rb
@@ -3,7 +3,13 @@ require_relative '../../lazar/lib/lazar'
 include OpenTox
 
 summary = [] 
-Validation::RepeatedCrossValidation.find(File.read(ARGV[0]).chomp).crossvalidations.each do |cv|
-  summary << cv.statistics
+#Validation::RepeatedCrossValidation.find(File.read(ARGV[0]).chomp).crossvalidations.each do |cv|
+#  summary << cv.statistics
+#end
+cv = Validation::RepeatedCrossValidation.all.last.crossvalidations.each do |cv|
+#cv = Validation::RepeatedCrossValidation.find("5bbb9151ca626916248b328b").crossvalidations.first
+p cv.id
+p cv.statistics
 end
-puts JSON.pretty_generate(summary)
+  
+#puts JSON.pretty_generate(summary)