compound.rb in lazar-0.0.9

- old
+ new

@@ -52,23 +52,21 @@
     File.unlink(testbild)
   end
 
   def test_inchikey
     c = OpenTox::Compound.from_inchi "InChI=1S/C6H6/c1-2-4-6-5-3-1/h1-6H"
-    p c
     assert_equal "UHOVQNZJYSORNB-UHFFFAOYSA-N", c.inchikey
   end
 
   def test_cid
     c = OpenTox::Compound.from_inchi "InChI=1S/C6H6/c1-2-4-6-5-3-1/h1-6H"
     assert_equal "241", c.cid
   end
 
   def test_chemblid
     c = OpenTox::Compound.from_inchi "InChI=1S/C6H6/c1-2-4-6-5-3-1/h1-6H"
-    #assert_equal "CHEMBL277500", c.chemblid
-    assert_equal "CHEMBL581676", c.chemblid
+    assert_equal "CHEMBL277500", c.chemblid
   end
 
   def test_sdf_storage
     c = OpenTox::Compound.from_smiles "CC(=O)CC(C)C#N"
     c.sdf
@@ -76,25 +74,136 @@
   end
 
   def test_fingerprint
     c = OpenTox::Compound.from_smiles "CC(=O)CC(C)C#N"
 
-    assert c.fp4.collect{|fid| Feature.find(fid).name}.include? ("1,3-Tautomerizable")
-    assert_equal c.fp4.size, c.fp4_size
+    assert_equal 9, c.fingerprint("FP4").size
   end
 
   def test_neighbors
     d = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.csv")
     d.compounds.each do |c|
-      refute_nil c.fp4
+      refute_nil c.fingerprint("MP2D")
     end
     c = d.compounds[371]
-    assert c.neighbors.size >= 19
+    n = c.fingerprint_neighbors({:type => "FP4", :min_sim => 0.7, :training_dataset_id => d.id })
+    assert n.size >= 18, "Neighbors size (#{n.size}) should be larger than 17"
   end
 
   def test_openbabel_segfault
     inchi = "InChI=1S/C19H27NO7/c1-11-9-19(12(2)27-19)17(23)26-14-6-8-20(4)7-5-13(15(14)21)10-25-16(22)18(11,3)24/h5,11-12,14,24H,6-10H2,1-4H3/b13-5-/t11-,12-,14-,18-,19?/m1/s1"
 
     c = Compound.from_inchi(inchi)
     assert_equal inchi, c.inchi
+  end
+
+  def test_openbabel_fingerprint
+    [
+      "CC(=O)CC(C)C#N",
+      "CC(=O)CC(C)C",
+      "C(=O)CC(C)C#N",
+    ].each do |smi|
+      c = OpenTox::Compound.from_smiles smi
+      refute_nil c.fingerprint("FP4")
+    end
+  end
+
+  def test_fingerprint_neighbors
+    types = ["FP2", "FP3", "FP4", "MACCS"]
+    min_sim = 0.7
+    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.csv")
+    [
+      "CC(=O)CC(C)C#N",
+      "CC(=O)CC(C)C",
+      "C(=O)CC(C)C#N",
+    ].each do |smi|
+      c = OpenTox::Compound.from_smiles smi
+      types.each do |type|
+        neighbors = c.fingerprint_neighbors({:type => type, :training_dataset_id => training_dataset.id, :min_sim => min_sim})
+        unless type == "FP2" and smi == "CC(=O)CC(C)C#N" or smi == "C(=O)CC(C)C#N" and (type == "FP2" or type == "MACCS")
+          refute_empty neighbors
+        end
+      end
+    end
+  end
+
+  def test_mna
+    c = OpenTox::Compound.from_smiles "N#[N+]C1=CC=CC=C1.F[B-](F)(F)F"
+    assert_equal 18, c.fingerprint("MNA").size
+    assert_equal 9, c.fingerprint("MNA").uniq.size
+  end
+
+  def test_mpd
+    c = OpenTox::Compound.from_smiles "N#[N+]C1=CC=CC=C1.F[B-](F)(F)F"
+    assert 13, c.fingerprint("MP2D").size
+    assert 7, c.fingerprint("MP2D").uniq.size
+  end
+
+  def test_fingerprint_count_neighbors
+    types = ["MP2D", "MNA"]
+    min_sim = 0.0
+    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.csv")
+    [
+      "CC(=O)CC(C)C#N",
+      "CC(=O)CC(C)C",
+      "C(=O)CC(C)C#N",
+    ].each do |smi|
+      c = OpenTox::Compound.from_smiles smi
+      types.each do |type|
+        neighbors = c.fingerprint_count_neighbors({:type => type, :training_dataset_id => training_dataset.id, :min_sim => min_sim})
+        if type == "FP4"
+          fp4_neighbors = c.neighbors
+          neighbors.each do |n|
+            assert_includes fp4_neighbors, n
+          end
+        end
+      end
+    end
+  end
+
+  def test_fingerprint_db_neighbors
+    #skip
+    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.csv")
+    [
+      "CC(=O)CC(C)C#N",
+      "CC(=O)CC(C)C",
+      "C(=O)CC(C)C#N",
+    ].each do |smi|
+      c = OpenTox::Compound.from_smiles smi
+      t = Time.now
+      neighbors = c.db_neighbors(:training_dataset_id => training_dataset.id, :min_sim => 0.2)
+      p Time.now - t
+      t = Time.now
+      neighbors2 = c.fingerprint_neighbors({:type => "MP2D", :training_dataset_id => training_dataset.id, :min_sim => 0.2})
+      p Time.now - t
+      p neighbors.size
+      p neighbors2.size
+      #p neighbors
+      #p neighbors2
+      #p neighbors2 - neighbors
+      #assert_equal neighbors, neighbors2
+    end
+  end
+
+  def test_molecular_weight
+    c = OpenTox::Compound.from_smiles "CC(=O)CC(C)C"
+    assert_equal 100.15888, c.molecular_weight
+  end
+
+  def test_mg_conversions
+    # TODO fix!
+    skip
+    c = OpenTox::Compound.from_smiles "O"
+    mw = c.molecular_weight
+    assert_equal 18.01528, mw
+    assert_equal 0.8105107141417474, c.logmmol_to_mg(4.34688225631145, mw)
+    assert_equal 9007.64, c.mmol_to_mg(500, mw)
+    assert_equal 2437.9999984148976, c.logmg_to_mg(3.387033701)
+  end
+
+  def test_physchem
+    c = OpenTox::Compound.from_smiles "CC(=O)CC(C)C"
+    assert_equal PhysChem::OBDESCRIPTORS.size, c.physchem.size
+    assert_equal PhysChem::OBDESCRIPTORS.size, c.physchem(PhysChem.openbabel_descriptors).size
+    assert_equal PhysChem::unique_descriptors.size, c.physchem(PhysChem.unique_descriptors).size
   end
 end