rdworks 0.37.1__tar.gz → 0.39.1__tar.gz

{rdworks-0.37.1 → rdworks-0.39.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rdworks
-Version: 0.37.1
+Version: 0.39.1
 Summary: Frequently used tasks built on RDKit and other tools
 Author-email: Sung-Hun Bae <sunghun.bae@gmail.com>
 Maintainer-email: Sung-Hun Bae <sunghun.bae@gmail.com>

{rdworks-0.37.1 → rdworks-0.39.1}/src/rdworks/__init__.py

@@ -1,4 +1,4 @@
-__version__ = '0.37.1'
+__version__ = '0.39.1'
 
 from rdworks.conf    import Conf
 from rdworks.mol     import Mol

{rdworks-0.37.1 → rdworks-0.39.1}/src/rdworks/conf.py

@@ -39,11 +39,20 @@ class Conf:
         """
         assert isinstance(molecule, str | Chem.Mol) or molecule is None
 
-        self.rdmol = None # must contain one and only one rdkit conformer
         self.name = name
+        self.rdmol = None   # must contain one and only one rdkit conformer
         self.natoms = 0
-        self.charge = 0 # molecular charge
-        self.spin = 1 # molecular spin multiplicity for ASE
+        self.charge = 0     # molecular formal charge
+        self.spin = 1       # molecular spin multiplicity for ASE
+        # Molecular spin multiplicity describes the number of possible orientations of 
+        # spin angular momentum for a given molecule, essentially indicating the total 
+        # number of unpaired electrons.
+        # Spin Angular Momentum (S): up (+1/2) or down (-1/2)
+        # Spin Multiplicity (2S + 1)
+        # 0 unpaired electron  has S = 0,   2S + 1 = 0, called a singlet. 
+        # 1 unpaired electron  has S = 1/2, 2S + 1 = 2, called a doublet (radical). 
+        # 2 unpaired electrons has S = 1,   2S + 1 = 3, called a triplet.  
+        
         self.props = {}
 
         if molecule is None:
@@ -63,10 +72,12 @@ class Conf:
 
         num_atoms = self.rdmol.GetNumAtoms()
         tot_atoms = self.rdmol.GetNumAtoms(onlyExplicit=False)
+        
         assert num_atoms == tot_atoms, "Conf() Error: missing hydrogens"
+        
         self.natoms = num_atoms
         self.charge = rdmolops.GetFormalCharge(self.rdmol)
-        self.props.update({'atoms': num_atoms})
+        self.props.update({'atoms': self.natoms, 'charge': self.charge})
      
         assert self.rdmol.GetConformer().Is3D(), "Conf() Error: not 3D"
 
@@ -148,7 +159,6 @@ class Conf:
 
         return self
     
-
     
     def optimize(self, 
                  calculator: str | Callable = 'MMFF94',
@@ -402,6 +412,95 @@ class Conf:
         return {i: d[:4] for i, d in enumerate(get_torsion_atoms(self.rdmol, strict))}
     
 
+    def torsion_energies_one(self,
+                             calculator: str | Callable,
+                             indices: tuple,
+                             simplify: bool = True,
+                             fmax: float = 0.05,
+                             interval: float = 20.0,
+                             use_converged_only: bool = True,
+                             **kwargs) -> Self:
+        """Calculate potential energy profile for a torsion angle.
+
+        Args:
+            calculator (str | Callable): 'MMFF', 'UFF', or ASE calculator.
+            indices (tuple): atom indices (i,j,k,l) for a torsion angle
+            simplify (bool, optional): whether to use fragementation. Defaults to True.
+            fmax (float, optional): convergence limit for optimize. Defaults to 0.05.
+            interval (float, optional): angle intervals. Defaults to 20.0.
+            use_converged_only (bool, optional): whether to use only converged data. Defaults to True.
+
+        Returns:
+            Self: modified self.
+        """
+        ref_conf = self.copy()
+
+        data = {'indices': indices,
+                'angle': [], 
+                'init': [], 
+                'last': [], 
+                'Converged': [],
+                }
+        
+        if simplify:
+            frag, frag_ijkl = create_torsion_fragment(ref_conf.rdmol, indices)
+            frag_conf = Conf(frag)
+            for angle in np.arange(-180.0, 180.0, interval): 
+                # Iterated numpy.ndarray does not contain the last 180: -180., ..., (180).
+                conf = frag_conf.copy()
+                conf.set_torsion(*frag_ijkl, angle) # atoms bonded to `l` move.
+                conf = conf.optimize(calculator, fmax, **kwargs)
+                # conf.optimize() updates coordinates and conf.props: 
+                #   `E_tot_init(kcal/mol)`, `E_tot(kcal/mol)`, `Converged`.
+                data['angle'].append(angle)
+                data['init'].append(conf.props['E_tot_init(kcal/mol)'])
+                data['last'].append(conf.props['E_tot(kcal/mol)'])
+                data['Converged'].append(conf.props['Converged'])
+                frag_cleaned, _ = clean_2d(frag, reset_isotope=True, remove_H=True)
+                # to serialize the molecule
+                data['frag'] = Chem.MolToMolBlock(frag_cleaned)
+                data['frag_indices'] = frag_ijkl
+        else:
+            for angle in np.arange(-180.0, 180.0, interval): 
+                # Iterated numpy.ndarray does not contain the last 180: -180., ..., (180).
+                conf = ref_conf.copy()
+                conf.set_torsion(*indices, angle) # atoms bonded to `l` move.
+                conf = conf.optimize(calculator, fmax, **kwargs)
+                # conf.optimize() updates coordinates and conf.props: 
+                #   `E_tot_init(kcal/mol)`, `E_tot(kcal/mol)`, `Converged`.
+                data['angle'].append(conf.props['angle'])
+                data['init'].append(conf.props['E_tot_init(kcal/mol)'])
+                data['last'].append(conf.props['E_tot(kcal/mol)'])
+                data['Converged'].append(conf.props['Converged'])
+        
+        # Post-processing        
+        if use_converged_only:
+            data['angle'] = list(itertools.compress(data['angle'], data['Converged']))
+            data['init' ] = list(itertools.compress(data['init' ], data['Converged']))
+            data['last' ] = list(itertools.compress(data['last' ], data['Converged']))
+        
+        relax = np.array(data['init']) - np.median(data['last'])
+        E_rel = relax - np.min(relax)
+
+        torsion_energy_profile = {
+            'indices' : data['indices'],
+            'angle' : np.round(np.array(data['angle']), 1).tolist(), # np.ndarray -> list for serialization
+            'E_rel(kcal/mol)': np.round(E_rel, 2).tolist(), # np.ndarray -> list for serialization
+            }
+        
+        if simplify:
+            torsion_energy_profile.update({
+                'frag' : data.get('frag', None),
+                'frag_indices' : data.get('frag_indices', None),
+            })
+
+        self.props['torsion'] = torsion_energy_profile
+        self.props['torsion_calculator'] = str(calculator)
+
+        return self
+    
+
+        
     def torsion_energies(self,
                          calculator: str | Callable,
                          torsion_key: int | None = None,
@@ -417,7 +516,7 @@ class Conf:
 
         Args:
             calculator (str | Callable): 'MMFF', 'UFF', or ASE calculator.
-            torsion_key (int | None): torsion index to calculate. Defaults to None (all).
+            torsion_key (int | None): key to the torsion indices to calculate. Defaults to None (all).
             simplify (bool, optional): whether to use fragment surrogate. Defaults to True.
             fmax (float, optional): fmax of ASE optimizer. Defaults to 0.05.
             interval (float, optional): interval of torsion angles in degree. Defaults to 15.0.
@@ -427,80 +526,23 @@ class Conf:
             Self: modified self.
         """
 
-        if torsion_key is None:
-            torsion_atoms_indices = self.torsion_atoms()
-            # {0: (5, 4, 3, 1)}
-        else:
-            torsion_atoms_indices = {torsion_key: self.torsion_atoms()[torsion_key]}
+        torsion_atoms_dict = self.torsion_atoms() # {0: (5, 4, 3, 1)}
+        if (torsion_key is not None) and torsion_atoms_dict.get(torsion_key):
+            torsion_atoms_dict = {torsion_key: torsion_atoms_dict.get(torsion_key)}
+            # single torsion angle atom indices
 
-        ref_conf = self.copy()
-
-        data = {}
-
-        if simplify:
-            for tk, indices in torsion_atoms_indices.items():
-                frag, frag_ijkl = create_torsion_fragment(ref_conf.rdmol, indices)
-                frag_conf = Conf(frag)
-                data[tk] = {'indices': indices, 'angle':[], 'init':[], 'last':[], 'Converged':[]}
-                for angle in np.arange(-180.0, 180.0, interval): 
-                    # Iterated numpy.ndarray does not contain the last 180: -180., ..., (180).
-                    conf = frag_conf.copy()
-                    conf.props.update({'torsion_key': tk, 'angle': float(angle)})
-                    conf.set_torsion(*frag_ijkl, angle) # atoms bonded to `l` move.
-                    conf = conf.optimize(calculator, fmax, **kwargs)
-                    # conf.optimize() updates coordinates and conf.props: 
-                    #   `angle`, `E_tot_init(kcal/mol)`, `E_tot(kcal/mol)`, `Converged`.
-                    tk = conf.props['torsion_key']
-                    data[tk]['angle'].append(conf.props['angle'])
-                    data[tk]['init'].append(conf.props['E_tot_init(kcal/mol)'])
-                    data[tk]['last'].append(conf.props['E_tot(kcal/mol)'])
-                    data[tk]['Converged'].append(conf.props['Converged'])
-                    frag_cleaned, _ = clean_2d(frag, reset_isotope=True, remove_H=True)
-                    # to serialize the molecule
-                    data[tk]['frag'] = Chem.MolToMolBlock(frag_cleaned)
-                    data[tk]['frag_indices'] = frag_ijkl
+        conf = self.copy()
 
-        else:
-            # mol.confs will be populated with torsion conformers. 
-            # It is designed for a batch optimization in the future.
-            torsion_angle_confs = []
-            for tk, indices in torsion_atoms_indices.items():
-                data[tk] = {'indices': indices, 'angle':[], 'init':[], 'last':[], 'Converged':[]}
-                for angle in np.arange(-180.0, 180.0, interval): 
-                    # Iterated numpy.ndarray does not contain the last 180: -180., ..., (180).
-                    x = ref_conf.copy()
-                    x.props.update({'torsion_key': tk, 'angle': float(angle)})
-                    x.set_torsion(*indices, angle) # atoms bonded to `l` move.
-                    torsion_angle_confs.append(x)
-
-            # Calculate relaxation energies
-            for conf in torsion_angle_confs:
-                conf = conf.optimize(calculator, fmax, **kwargs)
-                # conf.optimize() updates coordinates and conf.props: 
-                #   `angle`, `E_tot_init(kcal/mol)`, `E_tot(kcal/mol)`, `Converged`.
-                tk = conf.props['torsion_key']
-                data[tk]['angle'].append(conf.props['angle'])
-                data[tk]['init'].append(conf.props['E_tot_init(kcal/mol)'])
-                data[tk]['last'].append(conf.props['E_tot(kcal/mol)'])
-                data[tk]['Converged'].append(conf.props['Converged'])
-        
-        # Post-processing
         torsion_energy_profiles = {}
-        for tk, dictdata in data.items():
-            if use_converged_only:
-                dictdata['angle'] = list(itertools.compress(dictdata['angle'], dictdata['Converged']))
-                dictdata['init'] = list(itertools.compress(dictdata['init'], dictdata['Converged']))
-                dictdata['last'] = list(itertools.compress(dictdata['last'], dictdata['Converged']))
-            relax = np.array(dictdata['init']) - np.median(dictdata['last'])
-            E_rel = relax - np.min(relax)
-            torsion_energy_profiles[tk] = {
-                'indices' : dictdata['indices'],
-                'angle' : np.round(np.array(dictdata['angle']), 1).tolist(), # np.ndarray -> list for serialization
-                'E_rel(kcal/mol)': np.round(E_rel, 2).tolist(), # np.ndarray -> list for serialization
-                'frag' : dictdata.get('frag', None),
-                'frag_indices' : dictdata.get('frag_indices', None),
-                }
-
+        for tk, indices in torsion_atoms_dict.items():
+            conf = conf.torsion_energies_one(calculator, 
+                                             indices, 
+                                             simplify, 
+                                             fmax, 
+                                             interval, 
+                                             use_converged_only)
+            torsion_energy_profiles[tk] = conf.props['torsion']
+        
         self.props['torsion'] = torsion_energy_profiles
         self.props['torsion_calculator'] = str(calculator)
 
@@ -548,8 +590,10 @@ class Conf:
             str: serialized string for json.loads()
         """
         serialized = json.dumps({
-            'name' : self.name, 
-            'natoms': self.natoms, 
+            'name' : self.name,
+            'natoms': self.natoms,
+            'charge': self.charge,
+            'spin': self.spin,
             'props' : recursive_round(self.props, decimals),
             'molblock' : self.to_molblock(),
             })
@@ -573,7 +617,9 @@ class Conf:
         data = json.loads(serialized)
 
         self.name = data['name']
-        self.natoms = data['natoms']
+        self.natoms = int(data['natoms'])
+        self.charge = int(data['charge'])
+        self.spin = int(data['spin'])
         self.props = data['props']
         self.rdmol = Chem.MolFromMolBlock(data['molblock'], sanitize=False, removeHs=False)
 

{rdworks-0.37.1 → rdworks-0.39.1}/src/rdworks/mol.py

@@ -33,7 +33,7 @@ from rdkit.ML.Cluster import Butina
 from PIL import Image
 
 from rdworks.conf import Conf
-from rdworks.std import desalt_smiles, standardize, clean_2d
+from rdworks.std import generate_inchi_key, desalt_smiles, standardize, clean_2d
 from rdworks.xml import list_predefined_xml, get_predefined_xml, parse_xml
 from rdworks.scaffold import rigid_fragment_indices
 from rdworks.descriptor import rd_descriptor, rd_descriptor_f
@@ -83,17 +83,18 @@ class Mol:
         """
         assert isinstance(molecule, str | Chem.Mol | Conf) or molecule is None
 
+        self.name = ''
         self.rdmol = None # 2D, one and only one Conformer
         self.smiles = '' # isomeric SMILES
         self.confs = [] # container for 3D conformers
-        self.name = ''
         self.InChIKey = '' # 27 characters (SHA-256 hash of InChI)
-        self.InChI = ''
         self.props = {}
-        self.fp = None
+        
         self.max_workers = max_workers
         self.chunksize = chunksize
         self.progress = progress
+        
+        self.fp = None
 
         if molecule is None:
             return
@@ -141,8 +142,7 @@ class Mol:
             self.name = 'untitled'
 
         self.rdmol.SetProp('_Name', self.name) # _Name can't be None
-        self.InChI = Chem.MolToInchi(self.rdmol)
-        self.InChIKey = inchi.InchiToInchiKey(self.InChI)
+        self.InChIKey = generate_inchi_key(self.rdmol)
         self.props.update({
             'aka' : [], # <-- to be set by MolLibr.unique()
             'atoms' : self.rdmol.GetNumAtoms(), # hydrogens not excluded?
@@ -1570,6 +1570,7 @@ class Mol:
         serialized = json.dumps({
             'name'  : self.name,
             'smiles': self.smiles,
+            'InChIKey': self.InChIKey,
             'props' : recursive_round(self.props, decimals),
             'confs' : [conf.serialize() for conf in self.confs],
             })
@@ -1596,8 +1597,7 @@ class Mol:
         self.smiles = data['smiles'] # isomeric SMILES, no H
         self.rdmol = Chem.MolFromSmiles(data['smiles']) # for 2D depiction
         self.rdmol.SetProp('_Name', self.name)
-        self.InChI = Chem.MolToInchi(self.rdmol)
-        self.InChIKey = inchi.InchiToInchiKey(self.InChI)
+        self.InChIKey = data['InChIKey']
         self.props = data['props']
         self.confs = [Conf().deserialize(_) for _ in data['confs']] # for 3D conformers (iterable)
         

{rdworks-0.37.1 → rdworks-0.39.1}/src/rdworks/std.py

@@ -1,10 +1,32 @@
 import operator
 
 from rdkit import Chem
-from rdkit.Chem import rdDepictor
+from rdkit.Chem import rdDepictor, inchi
 from rdkit.Chem.MolStandardize import rdMolStandardize
 
 
+def generate_inchi_key(rdmol: Chem.Mol) -> str:
+    """Generate InChIKey.
+
+    Note:
+        - An InChIKey is a 27-character string consisting of three parts: 
+            - 14 characters: Derived from the connectivity layer of the InChI.
+            - Hyphen (-): Separates the first two blocks.
+            - 9 characters: Derived from the remaining InChI layers.
+            - A hyphen (-): Separates the second and third blocks.
+            - A final checksum character: Ensures the integrity of the key.
+
+    Args:
+        rdmol (Chem.Mol): input molecule
+
+    Returns:
+        str: 27-character InChIKey
+    """
+    InChI = Chem.MolToInchi(rdmol)
+    InChIKey = inchi.InchiToInchiKey(InChI)
+    
+    return InChIKey
+
 
 def desalt_smiles(smiles: str) -> tuple[str, Chem.Mol]:
     """Remove salt(s) from SMILES.

{rdworks-0.37.1 → rdworks-0.39.1}/src/rdworks.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rdworks
-Version: 0.37.1
+Version: 0.39.1
 Summary: Frequently used tasks built on RDKit and other tools
 Author-email: Sung-Hun Bae <sunghun.bae@gmail.com>
 Maintainer-email: Sung-Hun Bae <sunghun.bae@gmail.com>

{rdworks-0.37.1 → rdworks-0.39.1}/tests/test_basics.py

@@ -475,7 +475,6 @@ def test_torsion_fragment():
     frag, frag_ijkl = create_torsion_fragment(mol.confs[0].rdmol, ta[6])
     assert frag_ijkl == (5, 6, 7, 12)
 
-
     mol2 = Mol(molecule='CC(=O)Nc1ccc(O)cc1', name='acetaminophen.3').make_confs(n=1)
     ta2 = mol2.torsion_atoms()
     # {0: (5, 4, 3, 1)}
@@ -486,12 +485,41 @@ def test_torsion_fragment():
     assert frag_ijkl == ta2[0]
 
 
+def test_torsion_fragment_from_conf():
+    from rdworks.torsion import create_torsion_fragment
+    mol = Mol(molecule="CC(C)C1=C(C(=C(N1CC[C@H](C[C@H](CC(=O)O)O)O)C2=CC=C(C=C2)F)C3=CC=CC=C3)C(=O)NC4=CC=CC=C4",
+              name="atorvastatin").make_confs(n=1)
+    ref_conf = mol.confs[0]
+    ta = ref_conf.torsion_atoms()
+    assert len(ta) == 12
+    # {0: (0, 1, 3, 7),  1: (3, 4, 32, 33),  2: (4, 5, 26, 27),  3: (7, 6, 19, 20),
+    # 4: (3, 7, 8, 9),  5: (7, 8, 9, 10),   6: (8, 9, 10, 18),   7: (18, 10, 11, 12),
+    # 8: (10, 11, 12, 17),  9: (17, 12, 13, 14),   10: (12, 13, 14, 15),   11: (36, 35, 34, 32)}
+    frag, frag_ijkl = create_torsion_fragment(ref_conf.rdmol, ta[6])
+    assert frag_ijkl == (5, 6, 7, 12)
+
+    ref_conf = ref_conf.torsion_energies(calculator='MMFF94', torsion_key=6, interval=15)
+
+    mol2 = Mol(molecule='CC(=O)Nc1ccc(O)cc1', name='acetaminophen.3').make_confs(n=1)
+    ref_conf2 = mol2.confs[0]
+    ta2 = ref_conf2.torsion_atoms()
+    # {0: (5, 4, 3, 1)}
+    assert len(ta2) == 1
+    frag, frag_ijkl = create_torsion_fragment(ref_conf2.rdmol, ta2[0])
+    # expects no fragmentation
+    assert frag == ref_conf2.rdmol
+    assert frag_ijkl == ta2[0]
+
+    ref_conf2 = ref_conf2.torsion_energies(calculator='MMFF94', interval=15)
+
+
 def test_torsion_energies():
     libr = MolLibr(torsion_dataset_smiles, torsion_dataset_names)
     with open(workdir / 'test_torsion_energies.html', 'w') as f:
         for mol in libr[:1]:
             mol = mol.make_confs().drop_confs(similar=True, similar_rmsd=0.3).sort_confs().rename()
-            mol = mol.optimize_confs(calculator='MMFF94').torsion_energies(calculator='MMFF94', interval=15)
+            mol = mol.optimize_confs(calculator='MMFF94').torsion_energies(calculator='MMFF94', 
+                                                                           interval=15)
             f.write(mol.to_html())
             print(mol.dumps('torsion', decimals=2))