From 6d68a1ca94937a0553f61ebbbbd317dae54ce4e6 Mon Sep 17 00:00:00 2001
From: "helma@in-silico.ch" <helma@in-silico.ch>
Date: Mon, 12 Nov 2018 17:08:51 +0100
Subject: PubChem classification download

---
 test/use_cases.rb | 84 +++++++++++++++++++++++++++++++++++++++++++++----------
 1 file changed, 70 insertions(+), 14 deletions(-)

(limited to 'test/use_cases.rb')

diff --git a/test/use_cases.rb b/test/use_cases.rb
index 4842a18..59677cb 100644
--- a/test/use_cases.rb
+++ b/test/use_cases.rb
@@ -50,25 +50,81 @@ class UseCasesTest < MiniTest::Test
   end
 
   def test_public_models
-    skip
+    #skip
+    # TODO clean mongo
+    # PubChem Classification
+    [
+      {
+        :aid => 1205,
+        :species => "Rodents",
+        :endpoint => "Carcinogenicity",
+        :qmrf => {:group => "QMRF 4.12. Carcinogenicity", :name => "OECD 451 Carcinogenicity Studies"}
+      },{
+        :aid => 1208,
+        :species => "Rat",
+        :endpoint => "Carcinogenicity",
+        :qmrf => {:group => "QMRF 4.12. Carcinogenicity", :name => "OECD 451 Carcinogenicity Studies"}
+      },{
+        :aid => 1199,
+        :species => "Mouse",
+        :endpoint => "Carcinogenicity",
+        :qmrf => {:group => "QMRF 4.12. Carcinogenicity", :name => "OECD 451 Carcinogenicity Studies"}
+      }
+    ].each do |assay|
+      Download.pubchem_classification aid: assay[:aid], species: assay[:species], endpoint: assay[:endpoint], active: "carcinogen", inactive: "non-carcinogen", qmrf: qmrf
+    end
+
 =begin
-    #classification
-    aids = [
-      1205, #Rodents (multiple species/sites)
-      1208, # rat carc
-      1199 # mouse
-      # Mutagenicity
+    # Mutagenicity
+    kazius = Dataset.from_sdf_file "#{DATA_DIR}/cas_4337.sdf"
+    hansen = Dataset.from_csv_file "#{DATA_DIR}/hansen.csv"
+    efsa = Dataset.from_csv_file "#{DATA_DIR}/efsa.csv"
+    datasets = [kazius,hansen,efsa]
+    map = {"1" => "mutagen", "0" => "nonmutagen"}
+    training_dataset = Dataset.merge datasets: datasets, features: datasets.collect{|d| d.bioactivity_features.first}, value_maps: [nil,map,map], keep_original_features: false, remove_duplicates: true
+    # rename merged feature
+    training_dataset.merged_features.first.name = "Mutagenicity"
+    training_dataset.merged_features.first.save
+    Model::Validation.from_dataset training_dataset: training_dataset, prediction_feature: training_dataset.merged_features.first, species: "Salmonella typhimurium", endpoint: "Mutagenicity"
+    
+    # Blood Brain Barrier Penetration
+    bbb = Dataset.from_csv_file "#{DATA_DIR}/bbb.csv"
+    Model::Validation.from_dataset training_dataset: bbb, prediction_feature: bbb.bioactivity_features.first, species: "Human", endpoint: "Blood Brain Barrier Penetration"
 
+    # PubChem Regression
+    # TODO transformations
+    [
+      {
+        :aid => 1195,
+        :species => "Human",
+        :endpoint => "Maximum Recommended Daily Dose"
+      },{
+        :aid => 1208,
+        :species => "Rat (TD50)",
+        :endpoint => "Carcinogenicity"
+      },{
+        :aid => 1199,
+        :species => "Mouse (TD50)",
+        :endpoint => "Carcinogenicity"
+      },{
+        :aid => 1188,
+        :species => "Fathead minnow",
+        :endpoint => "Acute Toxicity"
+      }
+    ].each do |assay|
+      Model::Validation.from_pubchem_aid aid: assay[:aid], species: assay[;species], endpoint: assay[:endpoint], regression:true
+    end
 
-      1195 #MRDD
-      1188 #FHM
-      1208, # rat carc td50
-      1199 # mouse td50
-    
     # daphnia
-    # Blood Brain Barrier Penetration
-    # Lowest observed adverse effect level (LOAEL)
+    daphnia = Dataset.from_csv_file "#{DATA_DIR}/daphnia.csv"
+    Model::Validation.from_dataset training_dataset: daphnia, prediction_feature: daphnia.bioactivity_features.first, species: "Daphnia magna", endpoint: "Acute toxicity"
 
+    # LOAEL
+    loael = Dataset.from_csv_file "#{DATA_DIR}/loael.csv"
+    Model::Validation.from_dataset training_dataset: loael, prediction_feature: loael.bioactivity_features.first, species: "Rat", endpoint: "Lowest observed adverse effect level (LOAEL)"
+=end
+
+=begin
       # 1204  estrogen receptor
       # 1259408, # GENE-TOX
       # 1159563 HepG2 cytotoxicity assay
-- 
cgit v1.2.3