Update to rdkit md default

isicle/geometry.py

@@ -299,7 +299,7 @@ def __init__(self, **kwargs):
         self.__dict__.update(dict.fromkeys(self._defaults, self._default_value))
         self.__dict__.update(kwargs)
 
-    def _is_embedded(self):
+    def _is_embedded(self, mol):
         """
         Check if molecule has embedded 3D coordinates.
 
@@ -311,7 +311,7 @@ def _is_embedded(self):
         """
 
         try:
-            self.mol.GetConformers()
+            mol.GetConformer().Is3D()
             return True
         except:
             return False
@@ -386,17 +386,17 @@ def _forcefield_selector(forcefield, mw):
         # Embed molecule 3D coordinates
         if embed is True:
             # Attempt embedding
-            try:
-                Chem.AllChem.EmbedMolecule(mol)
-
-            # Use random coordinates
-            except:
-                Chem.AllChem.EmbedMolecule(mol, useRandomCoords=True)
+            res = Chem.AllChem.EmbedMolecule(mol)
+            if res == -1:
+                # Use random coordinates
+                res = Chem.AllChem.EmbedMolecule(mol, useRandomCoords=True)
+                if res == -1:
+                    raise ValueError("Embedding failure.")
 
         # Optimize according to supplied forcefield
         if forcefield is not None:
             # Check if embedded
-            if self._is_embedded():
+            if self._is_embedded(mol):
                 # Forcefield selection
                 if "mmff94s" in forcefield.lower():
                     _forcefield_selector(forcefield, mol)(

isicle/md.py

@@ -519,7 +519,7 @@ def set_geometry(self, geom):
         self.geom = geom
         self.basename = self.geom.basename
 
-    def configure(self, method: str = "distance", numConfs: int = 10, **kwargs):
+    def configure(self, method: str = "etkdgv3", numConfs: int = 10, **kwargs):
         """
         Set conformer generation parameters.
         Parameters
@@ -600,12 +600,14 @@ def _configure_etkdg2(self):
         """
         self.params = rdDistGeom.ETKDGv2()
 
-    def _configure_etkdg3(self):
+    def _configure_etkdg3(self, variant=True):
         """
         Set parameters for ETKDG conformer generation, based on work by Riniker and Landrum.
         Version 3: Updated sampling small rings AND macrocycles
         """
         self.params = rdDistGeom.ETKDGv3()
+        if variant is True:
+            self.params.useSmallRingTorsions = True
 
     def _configure_etkdg3_variant(self):
         """

tests/test_conformers.py

@@ -26,140 +26,137 @@ def random_energies():
 
 @pytest.fixture()
 def conformers():
-    x = [isicle.geometry.load(localfile('resources/geom_test.mol'))
-         for i in range(30)]
+    x = [isicle.geometry.load(localfile("resources/geom_test.mol")) for i in range(30)]
     return isicle.conformers.ConformationalEnsemble(x)
 
 
-@pytest.mark.parametrize('func,expected',
-                         [('boltzmann', isicle.conformers.boltzmann),
-                          ('simple', isicle.conformers.simple_average),
-                          ('lowest', isicle.conformers.lowest_energy),
-                          ('threshold', isicle.conformers.threshold)])
+@pytest.mark.parametrize(
+    "func,expected",
+    [
+        ("boltzmann", isicle.conformers.boltzmann),
+        ("simple", isicle.conformers.simple_average),
+        ("lowest", isicle.conformers.lowest_energy),
+        ("threshold", isicle.conformers.energy_threshold),
+    ],
+)
 def test__function_selector(func, expected):
     # Check correct class is yielded
     assert isicle.conformers._function_selector(func) == expected
 
 
-@pytest.mark.parametrize('func',
-                         [('boltzman'),
-                          ('smple'),
-                          ('lowest-energy'),
-                          ('thresh')])
+@pytest.mark.parametrize(
+    "func", [("boltzman"), ("smple"), ("lowest-energy"), ("thresh")]
+)
 def test__function_selector_fail(func):
     # Check unsupported selections
     with pytest.raises(ValueError):
         isicle.conformers._function_selector(func)
 
 
-@pytest.mark.parametrize('func,expected',
-                         [(isicle.conformers.boltzmann, True),
-                          (isicle.conformers.simple_average, False),
-                          (isicle.conformers.lowest_energy, True),
-                          (isicle.conformers.threshold, True)])
+@pytest.mark.parametrize(
+    "func,expected",
+    [
+        (isicle.conformers.boltzmann, True),
+        (isicle.conformers.simple_average, False),
+        (isicle.conformers.lowest_energy, True),
+        (isicle.conformers.energy_threshold, True),
+    ],
+)
 def test__energy_based(func, expected):
     assert isicle.conformers._energy_based(func) is expected
 
 
 def test_reduce(random_values, random_energies):
-    result = isicle.conformers.reduce(random_values,
-                                      func='boltzmann',
-                                      energy=random_energies,
-                                      index=None)
+    result = isicle.conformers.reduce(
+        random_values, func="boltzmann", energy=random_energies, index=None
+    )
 
-    assert abs(result['mean'] - 670.505) < 1E-3
-    assert abs(result['std'] - 26.566) < 1E-3
-    assert result['n'] == 30
+    assert abs(result["mean"] - 670.505) < 1e-3
+    assert abs(result["std"] - 26.566) < 1e-3
+    assert result["n"] == 30
 
 
-@pytest.mark.parametrize('index',
-                         [(None)])
+@pytest.mark.parametrize("index", [(None)])
 def test_boltzmann(random_values, random_energies, index):
-    result = isicle.conformers.boltzmann(random_values,
-                                         random_energies,
-                                         index=index)
+    result = isicle.conformers.boltzmann(random_values, random_energies, index=index)
 
-    assert abs(result['mean'] - 670.505) < 1E-3
-    assert abs(result['std'] - 26.566) < 1E-3
-    assert result['n'] == 30
+    assert abs(result["mean"] - 670.505) < 1e-3
+    assert abs(result["std"] - 26.566) < 1e-3
+    assert result["n"] == 30
 
 
-@pytest.mark.parametrize('index',
-                         [(None)])
+@pytest.mark.parametrize("index", [(None)])
 def test_simple_average(random_values, index):
-    result = isicle.conformers.simple_average(random_values,
-                                              index=index)
+    result = isicle.conformers.simple_average(random_values, index=index)
 
-    assert abs(result['mean'] - 451.660) < 1E-3
-    assert abs(result['std'] - 276.717) < 1E-3
-    assert result['n'] == 30
+    assert abs(result["mean"] - 451.660) < 1e-3
+    assert abs(result["std"] - 276.717) < 1e-3
+    assert result["n"] == 30
 
 
-@pytest.mark.parametrize('index',
-                         [(None)])
+@pytest.mark.parametrize("index", [(None)])
 def test_lowest_energy(random_values, random_energies, index):
-    result = isicle.conformers.lowest_energy(random_values,
-                                             random_energies,
-                                             index=index)
+    result = isicle.conformers.lowest_energy(
+        random_values, random_energies, index=index
+    )
 
-    assert abs(result['value'] - 667.237) < 1E-3
-    assert abs(result['energy'] - 0.016) < 1E-3
+    assert abs(result["value"] - 667.237) < 1e-3
+    assert abs(result["energy"] - 0.016) < 1e-3
 
 
-@pytest.mark.parametrize('index',
-                         [(None)])
+@pytest.mark.parametrize("index", [(None)])
 def test_threshold(random_values, random_energies, index):
-    result = isicle.conformers.threshold(random_values,
-                                         random_energies,
-                                         threshold=0.5,
-                                         index=index)
-
-    assert abs(result['mean'] - 414.716) < 1E-3
-    assert abs(result['std'] - 289.999) < 1E-3
-    assert result['n'] == 15
-
-
-@pytest.mark.parametrize('objects',
-                         [([Geometry(), Geometry(), Geometry()]),
-                          ([XYZGeometry(), XYZGeometry(), Geometry()])])
+    result = isicle.conformers.energy_threshold(
+        random_values, random_energies, threshold=0.5, index=index
+    )
+
+    assert abs(result["mean"] - 414.716) < 1e-3
+    assert abs(result["std"] - 289.999) < 1e-3
+    assert result["n"] == 15
+
+
+@pytest.mark.parametrize(
+    "objects",
+    [
+        ([Geometry(), Geometry(), Geometry()]),
+        ([XYZGeometry(), XYZGeometry(), Geometry()]),
+    ],
+)
 def test_build_conformational_ensemble(objects):
     isicle.conformers.build_conformational_ensemble(objects)
 
 
-@pytest.mark.parametrize('objects',
-                         [([Geometry(), 'abc', Geometry()]),
-                          ([XYZGeometry(), Geometry(), list()])])
+@pytest.mark.parametrize(
+    "objects",
+    [([Geometry(), "abc", Geometry()]), ([XYZGeometry(), Geometry(), list()])],
+)
 def test_build_conformational_ensemble_fail(objects):
     with pytest.raises(TypeError):
         isicle.conformers.build_conformational_ensemble(objects)
 
 
 class TestConformationalEnsemble:
-
-    @pytest.mark.parametrize('index',
-                             [(False)])
+    @pytest.mark.parametrize("index", [(False)])
     def test_reduce(self, conformers, random_values, random_energies, index):
         # Set values
         for c, v, e in zip(conformers, random_values, random_energies):
-            c.global_properties = {'dummy': v, 'energy': {'energy': [e]}}
+            c.global_properties = {"dummy": v, "energy": {"energy": [e]}}
 
         # Reduce attribute
-        result = conformers.reduce('dummy', func='boltzmann')
+        result = conformers.reduce("dummy", func="boltzmann")
 
         # Verify result
-        assert abs(result['mean'] - 670.505) < 1E-3
-        assert abs(result['std'] - 26.566) < 1E-3
-        assert result['n'] == 30
+        assert abs(result["mean"] - 670.505) < 1e-3
+        assert abs(result["std"] - 26.566) < 1e-3
+        assert result["n"] == 30
 
     def test__apply_method(self, conformers):
-        result = conformers._apply_method('get_natoms')
+        result = conformers._apply_method("get_natoms")
         assert all(x == 2 for x in result)
 
     def test__apply_function(self, conformers):
-        result = conformers._apply_function(isicle.qm.dft,
-                                            program='NWChem',
-                                            fmt='xyz')
+        result = conformers._apply_function(isicle.qm.dft, program="NWChem", fmt="xyz")
 
     def test_apply(self, conformers):
-        result = conformers.apply(method='get_natoms')
+        result = conformers.apply(method="get_natoms")
         assert all(x == 2 for x in result)