Merge pull request #41 from OBrink/explicite_hydrogen_inclusion

Explicite hydrogen inclusion

CITATION.cff

@@ -8,7 +8,7 @@ authors:
   given-names: "Kohulan"
   orcid: "https://orcid.org/0000-0003-1066-7792"
 title: "RanDepict"
-version: 1.1.7
+version: 1.1.8
 doi: 10.5281/zenodo.5205528
 date-released: 2021-08-17
 url: "https://github.com/OBrink/RanDepict"

RanDepict/__init__.py

@@ -21,7 +21,7 @@
 
 """
 
-__version__ = "1.1.7"
+__version__ = "1.1.8"
 
 __all__ = [
     "RanDepict",

RanDepict/cdk_functionalities.py

@@ -59,20 +59,29 @@ def cdk_depict(
         depiction = self._cdk_render_molecule(molecule, has_R_group, shape)
         return depiction
 
-    def _cdk_mol_block_to_cxsmiles(self, mol_block: str) -> str:
+    def _cdk_mol_block_to_cxsmiles(
+        self,
+        mol_block: str,
+        ignore_explicite_hydrogens: bool = True,
+    ) -> str:
         """
         This function takes a mol block str and returns the corresponding CXSMILES
         with coordinates using the CDK.
 
         Args:
             mol_block (str): mol block str
+            ignore_explicite_hydrogens (bool, optional): whether or not to ignore H
 
         Returns:
             str: CXSMILES
         """
         atom_container = self._cdk_mol_block_to_iatomcontainer(mol_block)
-        smi_gen = JClass("org.openscience.cdk.smiles.SmilesGenerator")
-        flavor = JClass("org.openscience.cdk.smiles.SmiFlavor")
+        if ignore_explicite_hydrogens:
+            cdk_base = "org.openscience.cdk."
+            manipulator = JClass(cdk_base + "tools.manipulator.AtomContainerManipulator")
+            atom_container = manipulator.copyAndSuppressedHydrogens(atom_container)
+        smi_gen = JClass(cdk_base + "smiles.SmilesGenerator")
+        flavor = JClass(cdk_base + "smiles.SmiFlavor")
         smi_gen = smi_gen(flavor.CxSmilesWithCoords)
         cxsmiles = smi_gen.create(atom_container)
         return cxsmiles
@@ -134,6 +143,68 @@ def _cdk_iatomcontainer_to_mol_block(self, i_atom_container) -> str:
         mol_str = string_writer.toString()
         return str(mol_str)
 
+    def _cdk_add_explicite_hydrogen_to_molblock(self, mol_block: str) -> str:
+        """
+        This function takes a mol block and returns the mol block with explicit
+        hydrogen atoms.
+
+        Args:
+            mol_block (str): mol block that describes a molecule
+
+        Returns:
+            str: The same mol block with explicit hydrogen atoms
+        """
+        i_atom_container = self._cdk_mol_block_to_iatomcontainer(mol_block)
+        cdk_base = "org.openscience.cdk."
+        manipulator = JClass(cdk_base + "tools.manipulator.AtomContainerManipulator")
+        manipulator.convertImplicitToExplicitHydrogens(i_atom_container)
+        mol_block = self._cdk_iatomcontainer_to_mol_block(i_atom_container)
+        return mol_block
+
+    def _cdk_add_explicite_hydrogen_to_smiles(self, smiles: str) -> str:
+        """
+        This function takes a SMILES str and uses CDK to add explicite hydrogen atoms.
+        It returns an adapted version of the SMILES str.
+
+        Args:
+            smiles (str): SMILES representation of a molecule
+
+        Returns:
+            smiles (str): SMILES representation of a molecule with explicite H
+        """
+        i_atom_container = self._cdk_smiles_to_IAtomContainer(smiles)
+        cdk_base = "org.openscience.cdk."
+        manipulator = JClass(cdk_base + "tools.manipulator.AtomContainerManipulator")
+        manipulator.convertImplicitToExplicitHydrogens(i_atom_container)
+        smi_flavor = JClass("org.openscience.cdk.smiles.SmiFlavor").Absolute
+        smiles_generator = JClass("org.openscience.cdk.smiles.SmilesGenerator")(
+            smi_flavor
+        )
+        smiles = smiles_generator.create(i_atom_container)
+        return str(smiles)
+
+    def _cdk_remove_explicite_hydrogen_from_smiles(self, smiles: str) -> str:
+        """
+        This function takes a SMILES str and uses CDK to remove explicite hydrogen atoms.
+        It returns an adapted version of the SMILES str.
+
+        Args:
+            smiles (str): SMILES representation of a molecule
+
+        Returns:
+            smiles (str): SMILES representation of a molecule with explicite H
+        """
+        i_atom_container = self._cdk_smiles_to_IAtomContainer(smiles)
+        cdk_base = "org.openscience.cdk."
+        manipulator = JClass(cdk_base + "tools.manipulator.AtomContainerManipulator")
+        i_atom_container = manipulator.copyAndSuppressedHydrogens(i_atom_container)
+        smi_flavor = JClass("org.openscience.cdk.smiles.SmiFlavor").Absolute
+        smiles_generator = JClass("org.openscience.cdk.smiles.SmilesGenerator")(
+            smi_flavor
+        )
+        smiles = smiles_generator.create(i_atom_container)
+        return str(smiles)
+
     def _cdk_get_depiction_generator(self, molecule, has_R_group: bool = False):
         """
         This function defines random rendering options for the structure

RanDepict/randepict.py

@@ -195,8 +195,13 @@ def random_depiction(
         Returns:
             np.array: Chemical structure depiction
         """
+        orig_styles = self._config.styles
+        # TODO: add this to depiction feature fingerprint
+        if self.random_choice([True] + [False] * 5):
+            smiles = self._cdk_add_explicite_hydrogen_to_smiles(smiles)
+            self._config.styles = [style for style in orig_styles if style != 'pikachu']
         depiction_functions = self.get_depiction_functions(smiles)
-
+        self._config.styles = orig_styles
         for _ in range(3):
             if len(depiction_functions) != 0:
                 # Pick random depiction function and call it
@@ -267,11 +272,17 @@ def random_depiction_with_coordinates(
         orig_styles = self._config.styles
         self._config.styles = [style for style in orig_styles if style != 'pikachu']
         depiction_functions = self.get_depiction_functions(smiles)
+        fun = self.random_choice(depiction_functions)
         self._config.styles = orig_styles
+        # TODO: add this to depiction feature fingerprint
+        if self.random_choice([True] + [False] * 5):
+            smiles = self._cdk_add_explicite_hydrogen_to_smiles(smiles)
         mol_block = self._smiles_to_mol_block(smiles,
-                                              self.random_choice(['rdkit', 'indigo', 'cdk']))
-        cxsmiles = self._cdk_mol_block_to_cxsmiles(mol_block)
-        fun = self.random_choice(depiction_functions)
+                                              self.random_choice(['rdkit',
+                                                                  'indigo',
+                                                                  'cdk']))
+        cxsmiles = self._cdk_mol_block_to_cxsmiles(mol_block,
+                                                   ignore_explicite_hydrogens=True)
         depiction = fun(mol_block=mol_block, shape=shape)
         if augment:
             depiction = self.add_augmentations(depiction)

RanDepict/random_markush_structure_generator.py

@@ -1,5 +1,3 @@
-from jpype import JClass
-# import sys
 from typing import List
 from .randepict import RandomDepictor
 
@@ -49,7 +47,7 @@ def insert_R_group_var(self, smiles: str, num: int) -> str:
         Returns:
             smiles (str): input SMILES with $num inserted R group variables
         """
-        smiles = self.add_explicite_hydrogen_to_smiles(smiles)
+        smiles = self.depictor._cdk_add_explicite_hydrogen_to_smiles(smiles)
         potential_replacement_positions = self.get_valid_replacement_positions(smiles)
         r_groups = []
         # Replace C or H in SMILES with *
@@ -66,7 +64,7 @@ def insert_R_group_var(self, smiles: str, num: int) -> str:
                 break
         # Remove explicite hydrogen again and get absolute SMILES
         smiles = "".join(smiles)
-        smiles = self.remove_explicite_hydrogen_from_smiles(smiles)
+        smiles = self.depictor._cdk_remove_explicite_hydrogen_from_smiles(smiles)
         # Replace * with R groups
         for r_group in r_groups:
             smiles = smiles.replace("*", r_group, 1)
@@ -136,53 +134,3 @@ def get_valid_replacement_positions(self, smiles: str) -> List[int]:
                 ]:
                     replacement_positions.append(index - 1)
         return replacement_positions
-
-    def add_explicite_hydrogen_to_smiles(self, smiles: str) -> str:
-        """
-        This function takes a SMILES str and uses CDK to add explicite hydrogen atoms.
-        It returns an adapted version of the SMILES str.
-
-        Args:
-            smiles (str): SMILES representation of a molecule
-
-        Returns:
-            smiles (str): SMILES representation of a molecule with explicite H
-        """
-        i_atom_container = self.depictor._cdk_smiles_to_IAtomContainer(smiles)
-
-        # Add explicite hydrogen atoms
-        cdk_base = "org.openscience.cdk."
-        manipulator = JClass(cdk_base + "tools.manipulator.AtomContainerManipulator")
-        manipulator.convertImplicitToExplicitHydrogens(i_atom_container)
-
-        # Create absolute SMILES
-        smi_flavor = JClass("org.openscience.cdk.smiles.SmiFlavor").Absolute
-        smiles_generator = JClass("org.openscience.cdk.smiles.SmilesGenerator")(
-            smi_flavor
-        )
-        smiles = smiles_generator.create(i_atom_container)
-        return str(smiles)
-
-    def remove_explicite_hydrogen_from_smiles(self, smiles: str) -> str:
-        """
-        This function takes a SMILES str and uses CDK to remove explicite hydrogen atoms.
-        It returns an adapted version of the SMILES str.
-
-        Args:
-            smiles (str): SMILES representation of a molecule
-
-        Returns:
-            smiles (str): SMILES representation of a molecule with explicite H
-        """
-        i_atom_container = self.depictor._cdk_smiles_to_IAtomContainer(smiles)
-        # Remove explicite hydrogen atoms
-        cdk_base = "org.openscience.cdk."
-        manipulator = JClass(cdk_base + "tools.manipulator.AtomContainerManipulator")
-        i_atom_container = manipulator.copyAndSuppressedHydrogens(i_atom_container)
-        # Create absolute SMILES
-        smi_flavor = JClass("org.openscience.cdk.smiles.SmiFlavor").Absolute
-        smiles_generator = JClass("org.openscience.cdk.smiles.SmilesGenerator")(
-            smi_flavor
-        )
-        smiles = smiles_generator.create(i_atom_container)
-        return str(smiles)

Tests/test_functions.py

@@ -412,6 +412,20 @@ def test_get_depiction_functions_normal(self):
         difference = set(observed) ^ set(expected)
         assert not difference
 
+    def test_add_explicite_hydrogen_to_smiles(self):
+        # Assert that hydrogen atoms are added
+        input_smiles = "CCC"
+        expected_output = "C([H])([H])([H])C([H])([H])C([H])([H])[H]"
+        observed_output = self.depictor._cdk_add_explicite_hydrogen_to_smiles(input_smiles)
+        assert expected_output == observed_output
+
+    def test_remove_explicite_hydrogen_to_smiles(self):
+        # Assert that hydrogen atoms are removed
+        input_smiles = "C([H])([H])([H])C([H])([H])C([H])([H])[H]"
+        expected_output = "CCC"
+        observed_output = self.depictor._cdk_remove_explicite_hydrogen_from_smiles(input_smiles)
+        assert expected_output == observed_output
+
     def test_get_depiction_functions_isotopes(self):
         # PIKAChU can't handle isotopes
         observed = self.depictor.get_depiction_functions("[13CH3]N1C=NC2=C1C(=O)N(C(=O)N2C)C")
@@ -516,19 +530,6 @@ def test_insert_R_group_var_can_be_depicted(self):
             depiction = self.depictor.random_depiction(output_smiles)
             assert type(depiction) == np.ndarray
 
-    def test_add_explicite_hydrogen_to_smiles(self):
-        # Assert that hydrogen atoms are added
-        input_smiles = "CCC"
-        expected_output = "C([H])([H])([H])C([H])([H])C([H])([H])[H]"
-        observed_output = self.markush_creator.add_explicite_hydrogen_to_smiles(input_smiles)
-        assert expected_output == observed_output
-
-    def test_remove_explicite_hydrogen_to_smiles(self):
-        # Assert that hydrogen atoms are removed
-        input_smiles = "C([H])([H])([H])C([H])([H])C([H])([H])[H]"
-        expected_output = "CCC"
-        observed_output = self.markush_creator.remove_explicite_hydrogen_from_smiles(input_smiles)
-        assert expected_output == observed_output
 
     def test_get_valid_replacement_positions_simple_chain(self):
         # Simple example case