Inchitomol (#86)

* fix molsanitize exception error catching

* linting

* isort on other stuff

* add inchitomol element

molpipeline/abstract_pipeline_elements/any2mol/string2mol.py

@@ -4,8 +4,11 @@
 
 import abc
 
-from molpipeline.abstract_pipeline_elements.core import AnyToMolPipelineElement
-from molpipeline.utils.molpipeline_types import OptionalMol
+from molpipeline.abstract_pipeline_elements.core import (
+    AnyToMolPipelineElement,
+    InvalidInstance,
+)
+from molpipeline.utils.molpipeline_types import OptionalMol, RDKitMol
 
 
 class StringToMolPipelineElement(AnyToMolPipelineElement, abc.ABC):
@@ -43,3 +46,60 @@ def pretransform_single(self, value: str) -> OptionalMol:
         OptionalMol
             RDKit molecule if representation was valid, else InvalidInstance.
         """
+
+
+class SimpleStringToMolElement(StringToMolPipelineElement, abc.ABC):
+    """Transforms string representation to RDKit Mol objects."""
+
+    def pretransform_single(self, value: str) -> OptionalMol:
+        """Transform string to molecule.
+
+        Parameters
+        ----------
+        value: str
+            string representation.
+
+        Returns
+        -------
+        OptionalMol
+            Rdkit molecule if valid string representation, else None.
+        """
+        if value is None:
+            return InvalidInstance(
+                self.uuid,
+                f"Invalid representation: {value}",
+                self.name,
+            )
+
+        if not isinstance(value, str):
+            return InvalidInstance(
+                self.uuid,
+                f"Not a string: {value}",
+                self.name,
+            )
+
+        mol: RDKitMol = self.string_to_mol(value)
+
+        if not mol:
+            return InvalidInstance(
+                self.uuid,
+                f"Invalid representation: {value}",
+                self.name,
+            )
+        mol.SetProp("identifier", value)
+        return mol
+
+    @abc.abstractmethod
+    def string_to_mol(self, value: str) -> RDKitMol:
+        """Transform string representation to molecule.
+
+        Parameters
+        ----------
+        value: str
+            string representation
+
+        Returns
+        -------
+        RDKitMol
+            Rdkit molecule if valid representation, else None.
+        """

molpipeline/any2mol/__init__.py

@@ -2,12 +2,14 @@
 
 from molpipeline.any2mol.auto2mol import AutoToMol
 from molpipeline.any2mol.bin2mol import BinaryToMol
+from molpipeline.any2mol.inchi2mol import InchiToMol
 from molpipeline.any2mol.sdf2mol import SDFToMol
 from molpipeline.any2mol.smiles2mol import SmilesToMol
 
 __all__ = [
     "AutoToMol",
     "BinaryToMol",
     "SmilesToMol",
+    "InchiToMol",
     "SDFToMol",
 ]

molpipeline/any2mol/auto2mol.py

@@ -9,6 +9,7 @@
     InvalidInstance,
 )
 from molpipeline.any2mol.bin2mol import BinaryToMol
+from molpipeline.any2mol.inchi2mol import InchiToMol
 from molpipeline.any2mol.sdf2mol import SDFToMol
 from molpipeline.any2mol.smiles2mol import SmilesToMol
 from molpipeline.utils.molpipeline_types import OptionalMol, RDKitMol
@@ -29,6 +30,7 @@ def __init__(
         uuid: Optional[str] = None,
         elements: tuple[AnyToMolPipelineElement, ...] = (
             SmilesToMol(),
+            InchiToMol(),
             BinaryToMol(),
             SDFToMol(),
         ),
@@ -44,7 +46,7 @@ def __init__(
         uuid: str, optional (default=None)
             Unique identifier of PipelineElement.
         elements: tuple[AnyToMol, ...], optional (default=(SmilesToMol(),
-            BinaryToMol(), SDFToMol()))
+            InchiToMol(), BinaryToMol(), SDFToMol()))
             Elements to try to transform the input to a molecule.
         """
         super().__init__(name=name, n_jobs=n_jobs, uuid=uuid)

molpipeline/any2mol/inchi2mol.py

@@ -0,0 +1,27 @@
+"""Classes ment to transform given inchi to a RDKit molecule."""
+
+from rdkit import Chem
+
+from molpipeline.abstract_pipeline_elements.any2mol.string2mol import (
+    SimpleStringToMolElement,
+)
+from molpipeline.utils.molpipeline_types import RDKitMol
+
+
+class InchiToMol(SimpleStringToMolElement):
+    """Transforms Inchi to RDKit Mol objects."""
+
+    def string_to_mol(self, value: str) -> RDKitMol:
+        """Transform Inchi string to molecule.
+
+        Parameters
+        ----------
+        value: str
+            Inchi string.
+
+        Returns
+        -------
+        RDKitMol
+            Rdkit molecule if valid Inchi, else None.
+        """
+        return Chem.MolFromInchi(value)

molpipeline/any2mol/smiles2mol.py

@@ -5,16 +5,15 @@
 from rdkit import Chem
 
 from molpipeline.abstract_pipeline_elements.any2mol.string2mol import (
-    StringToMolPipelineElement as _StringToMolPipelineElement,
+    SimpleStringToMolElement,
 )
-from molpipeline.abstract_pipeline_elements.core import InvalidInstance
-from molpipeline.utils.molpipeline_types import OptionalMol, RDKitMol
+from molpipeline.utils.molpipeline_types import RDKitMol
 
 
-class SmilesToMol(_StringToMolPipelineElement):
+class SmilesToMol(SimpleStringToMolElement):
     """Transforms Smiles to RDKit Mol objects."""
 
-    def pretransform_single(self, value: str) -> OptionalMol:
+    def string_to_mol(self, value: str) -> RDKitMol:
         """Transform Smiles string to molecule.
 
         Parameters
@@ -24,30 +23,7 @@ def pretransform_single(self, value: str) -> OptionalMol:
 
         Returns
         -------
-        OptionalMol
+        RDKitMol
             Rdkit molecule if valid SMILES, else None.
         """
-        if value is None:
-            return InvalidInstance(
-                self.uuid,
-                f"Invalid SMILES: {value}",
-                self.name,
-            )
-
-        if not isinstance(value, str):
-            return InvalidInstance(
-                self.uuid,
-                f"Not a string: {value}",
-                self.name,
-            )
-
-        mol: RDKitMol = Chem.MolFromSmiles(value)
-
-        if not mol:
-            return InvalidInstance(
-                self.uuid,
-                f"Invalid SMILES: {value}",
-                self.name,
-            )
-        mol.SetProp("identifier", value)
-        return mol
+        return Chem.MolFromSmiles(value)

tests/test_elements/test_any2mol/test_auto2mol.py

@@ -17,6 +17,9 @@
 SMILES_CL_BR = "NC(Cl)(Br)C(=O)O"
 SMILES_METAL_AU = "OC[C@H]1OC(S[Au])[C@H](O)[C@@H](O)[C@@H]1O"
 
+INCHI_BENZENE = "InChI=1S/C6H6/c1-2-4-6-5-3-1/h1-6H"
+INCHI_CHLOROBENZENE = "InChI=1S/C6H5Cl/c7-6-4-2-1-3-5-6/h1-5H"
+
 # SDF
 with gzip.open(TEST_DATA_DIR / "P86_B_400.sdf.gz") as file:
     SDF_P86_B_400 = file.read()
@@ -82,6 +85,31 @@ def test_auto2mol_for_smiles(self) -> None:
         )
         del log_block
 
+    def test_auto2mol_for_inchi(self) -> None:
+        """Test molecules can be read from inchi automatically."""
+
+        test_inchis = [INCHI_BENZENE, INCHI_CHLOROBENZENE]
+        expected_mols = [MOL_BENZENE, MOL_CHLOROBENZENE]
+
+        pipeline = Pipeline(
+            [
+                (
+                    "Auto2Mol",
+                    AutoToMol(),
+                ),
+            ]
+        )
+        log_block = rdBase.BlockLogs()
+        actual_mols = pipeline.fit_transform(test_inchis)
+        self.assertEqual(len(test_inchis), len(actual_mols))
+        self.assertTrue(
+            all(
+                Chem.MolToInchi(smiles_mol) == Chem.MolToInchi(original_mol)
+                for smiles_mol, original_mol in zip(actual_mols, expected_mols)
+            )
+        )
+        del log_block
+
     def test_auto2mol_for_sdf(self) -> None:
         """Test molecules can be read from sdf automatically."""