def estimateThermoViaGroupAdditivityForSaturatedStructWithoutSymmetryCorrection(self, molecule): """ Return the set of thermodynamic parameters corresponding to a given :class:`Molecule` object `molecule` by estimation using the group additivity values. If no group additivity values are loaded, a :class:`DatabaseError` is raised. The entropy is not corrected for the symmetry of the molecule. This should be done later by the calling function. """ assert not molecule.isRadical(), "This method is only for saturated non-radical species." # For thermo estimation we need the atoms to already be sorted because we # iterate over them; if the order changes during the iteration then we # will probably not visit the right atoms, and so will get the thermo wrong molecule.sortVertices() # Create the ThermoData object thermoData = ThermoData( Tdata = ([300,400,500,600,800,1000,1500],"K"), Cpdata = ([0.0,0.0,0.0,0.0,0.0,0.0,0.0],"J/(mol*K)"), H298 = (0.0,"kJ/mol"), S298 = (0.0,"J/(mol*K)"), ) cyclic = molecule.isCyclic() # Generate estimate of thermodynamics for atom in molecule.atoms: # Iterate over heavy (non-hydrogen) atoms if atom.isNonHydrogen(): # Get initial thermo estimate from main group database try: self.__addGroupThermoData(thermoData, self.groups['group'], molecule, {'*':atom}) except KeyError: logging.error("Couldn't find in main thermo database:") logging.error(molecule) logging.error(molecule.toAdjacencyList()) raise # Correct for gauche and 1,5- interactions if not cyclic: try: self.__addGroupThermoData(thermoData, self.groups['gauche'], molecule, {'*':atom}) except KeyError: pass try: self.__addGroupThermoData(thermoData, self.groups['int15'], molecule, {'*':atom}) except KeyError: pass try: self.__addGroupThermoData(thermoData, self.groups['other'], molecule, {'*':atom}) except KeyError: pass # Do ring corrections separately because we only want to match # each ring one time if cyclic: if molecule.getAllPolycyclicVertices(): # If the molecule has fused ring atoms, this implies that we are dealing # with a polycyclic ring system, for which separate ring strain corrections may not # be adequate. Therefore, we search the polycyclic thermo group corrections # instead of adding single ring strain corrections within the molecule. # For now, assume only one polycyclic RSC can be found per molecule try: self.__addGroupThermoData(thermoData, self.groups['polycyclic'], molecule, {}) except: logging.error("Couldn't find in polycyclic ring database:") logging.error(molecule) logging.error(molecule.toAdjacencyList()) raise else: rings = molecule.getSmallestSetOfSmallestRings() for ring in rings: # Make a temporary structure containing only the atoms in the ring # NB. if any of the ring corrections depend on ligands not in the ring, they will not be found! try: self.__addGroupThermoData(thermoData, self.groups['ring'], molecule, {}) except KeyError: logging.error("Couldn't find in ring database:") logging.error(ring) logging.error(ring.toAdjacencyList()) raise return thermoData
def estimateThermoViaGroupAdditivity(self, molecule): """ Return the set of thermodynamic parameters corresponding to a given :class:`Molecule` object `molecule` by estimation using the group additivity values. If no group additivity values are loaded, a :class:`DatabaseError` is raised. """ # For thermo estimation we need the atoms to already be sorted because we # iterate over them; if the order changes during the iteration then we # will probably not visit the right atoms, and so will get the thermo wrong molecule.sortVertices() # Create the ThermoData object thermoData = ThermoData( Tdata = ([300,400,500,600,800,1000,1500],"K"), Cpdata = ([0.0,0.0,0.0,0.0,0.0,0.0,0.0],"J/(mol*K)"), H298 = (0.0,"kJ/mol"), S298 = (0.0,"J/(mol*K)"), ) if molecule.getRadicalCount() > 0: # radical species return self.estimateRadicalThermoViaHBI(molecule, self.estimateThermoViaGroupAdditivity ) else: # non-radical species cyclic = molecule.isCyclic() # Generate estimate of thermodynamics for atom in molecule.atoms: # Iterate over heavy (non-hydrogen) atoms if atom.isNonHydrogen(): # Get initial thermo estimate from main group database try: self.__addGroupThermoData(thermoData, self.groups['group'], molecule, {'*':atom}) except KeyError: logging.error("Couldn't find in main thermo database:") logging.error(molecule) logging.error(molecule.toAdjacencyList()) raise # Correct for gauche and 1,5- interactions if not cyclic: try: self.__addGroupThermoData(thermoData, self.groups['gauche'], molecule, {'*':atom}) except KeyError: pass try: self.__addGroupThermoData(thermoData, self.groups['int15'], molecule, {'*':atom}) except KeyError: pass try: self.__addGroupThermoData(thermoData, self.groups['other'], molecule, {'*':atom}) except KeyError: pass # Do ring corrections separately because we only want to match # each ring one time if cyclic: if molecule.getAllPolycyclicVertices(): # If the molecule has fused ring atoms, this implies that we are dealing # with a polycyclic ring system, for which separate ring strain corrections may not # be adequate. Therefore, we search the polycyclic thermo group corrections # instead of adding single ring strain corrections within the molecule. # For now, assume only one polycyclic RSC can be found per molecule try: self.__addGroupThermoData(thermoData, self.groups['polycyclic'], molecule, {}) except: logging.error("Couldn't find in polycyclic ring database:") logging.error(molecule) logging.error(molecule.toAdjacencyList()) raise else: rings = molecule.getSmallestSetOfSmallestRings() for ring in rings: # Make a temporary structure containing only the atoms in the ring # NB. if any of the ring corrections depend on ligands not in the ring, they will not be found! try: self.__addGroupThermoData(thermoData, self.groups['ring'], molecule, {}) except KeyError: logging.error("Couldn't find in ring database:") logging.error(ring) logging.error(ring.toAdjacencyList()) raise # Correct entropy for symmetry number molecule.calculateSymmetryNumber() thermoData.S298.value_si -= constants.R * math.log(molecule.symmetryNumber) return thermoData
def calculateCyclicSymmetryNumber(molecule): """ Get the symmetry number correction for cyclic regions of a molecule. For complicated fused rings the smallest set of smallest rings is used. """ from rdkit.Chem.rdmolops import SanitizeMol from rdkit.Chem.rdchem import Mol symmetryNumber = 1 rings = molecule.getSmallestSetOfSmallestRings() # Get symmetry number for each ring in structure for ring0 in rings: # Make another copy structure structure = molecule.copy(True) ring = [structure.atoms[molecule.atoms.index(atom)] for atom in ring0] # Remove bonds of ring from structure for i, atom1 in enumerate(ring): for atom2 in ring[i + 1 :]: if structure.hasBond(atom1, atom2): structure.removeBond(atom1.edges[atom2]) structures = structure.split() groups = [] for struct in structures: for atom in ring: if struct.hasAtom(atom): struct.removeAtom(atom) groups.append(struct.split()) # Find equivalent functional groups on ring equivalentGroups = [] equivalentGroupCount = [] for group in groups: found = False for i, eqGroup in enumerate(equivalentGroups): if not found and len(group) == len(eqGroup): for g, eg in zip(group, eqGroup): if not g.isIsomorphic(eg): # The groups do not match break else: # The groups match found = True if found: # We've found a matching group, so increment its count equivalentGroupCount[i] += 1 break else: # No matching group found, so add it as a new group equivalentGroups.append(group) equivalentGroupCount.append(1) # Find equivalent bonds on ring equivalentBonds = [] for i, atom1 in enumerate(ring0): for atom2 in ring0[i + 1 :]: if molecule.hasBond(atom1, atom2): bond = molecule.getBond(atom1, atom2) found = False for eqBond in equivalentBonds: if not found: if bond.equivalent(eqBond[0]): eqBond.append(group) found = True if not found: equivalentBonds.append([bond]) # Find maximum number of equivalent groups and bonds minEquivalentGroups = min(equivalentGroupCount) maxEquivalentGroups = max(equivalentGroupCount) minEquivalentBonds = None maxEquivalentBonds = 0 for bonds in equivalentBonds: N = len(bonds) if minEquivalentBonds is None or N < minEquivalentBonds: minEquivalentBonds = N if N > maxEquivalentBonds: maxEquivalentBonds = N if maxEquivalentGroups == maxEquivalentBonds == len(ring): symmetryNumber *= len(ring) * 2 else: symmetryNumber *= min(minEquivalentGroups, minEquivalentBonds) return symmetryNumber
def calculateCyclicSymmetryNumber(molecule): """ Get the symmetry number correction for cyclic regions of a molecule. For complicated fused rings the smallest set of smallest rings is used. """ from rdkit.Chem.rdmolops import SanitizeMol from rdkit.Chem.rdchem import Mol mcopy = molecule.toRDKitMol(removeHs=True, returnMapping=False) SanitizeMol(mcopy) symmetryNumber = 1 # Get symmetry number for each ring in structure rings = molecule.getSmallestSetOfSmallestRings() for ring0 in rings: # Make copy of structure structure = molecule.copy(True) ring = [structure.atoms[molecule.atoms.index(atom)] for atom in ring0] # Figure out which atoms and bonds are aromatic and reassign appropriately: for i, atom1 in enumerate(ring0): for atom2 in ring0[i+1:]: if molecule.hasBond(atom1, atom2): if mcopy.GetBondBetweenAtoms(i,i+1) is not None: if str(mcopy.GetBondBetweenAtoms(i,i+1).GetBondType()) == 'AROMATIC': bond = molecule.getBond(atom1, atom2) bond.applyAction(['CHANGE_BOND', atom1, 'B', atom2]) atom1.atomType = atom2.atomType = rmgpy.molecule.atomTypes['Cb'] else: pass # Remove bonds of ring from structure for i, atom1 in enumerate(ring): for atom2 in ring[i+1:]: if structure.hasBond(atom1, atom2): structure.removeBond(atom1.edges[atom2]) structures = structure.split() groups = [] for struct in structures: for atom in ring: if struct.hasAtom(atom): struct.removeAtom(atom) groups.append(struct.split()) # Find equivalent functional groups on ring equivalentGroups = []; equivalentGroupCount = [] for group in groups: found = False for i, eqGroup in enumerate(equivalentGroups): if not found and len(group) == len(eqGroup): for g, eg in zip(group, eqGroup): if not g.isIsomorphic(eg): # The groups do not match break else: # The groups match found = True if found: # We've found a matching group, so increment its count equivalentGroupCount[i] += 1 break else: # No matching group found, so add it as a new group equivalentGroups.append(group) equivalentGroupCount.append(1) # Find equivalent bonds on ring equivalentBonds = [] for i, atom1 in enumerate(ring0): for atom2 in ring0[i+1:]: if molecule.hasBond(atom1, atom2): bond = molecule.getBond(atom1, atom2) found = False for eqBond in equivalentBonds: if not found: if bond.equivalent(eqBond[0]): eqBond.append(group) found = True if not found: equivalentBonds.append([bond]) # Find maximum number of equivalent groups and bonds minEquivalentGroups = min(equivalentGroupCount) maxEquivalentGroups = max(equivalentGroupCount) minEquivalentBonds = None maxEquivalentBonds = 0 for bonds in equivalentBonds: N = len(bonds) if minEquivalentBonds is None or N < minEquivalentBonds: minEquivalentBonds = N if N > maxEquivalentBonds: maxEquivalentBonds = N if maxEquivalentGroups == maxEquivalentBonds == len(ring): symmetryNumber *= len(ring) * 2 else: symmetryNumber *= min(minEquivalentGroups, minEquivalentBonds) #print len(ring), minEquivalentGroups, maxEquivalentGroups, minEquivalentBonds, maxEquivalentBonds, symmetryNumber return symmetryNumber
def calculateCyclicSymmetryNumber(molecule): """ Get the symmetry number correction for cyclic regions of a molecule. For complicated fused rings the smallest set of smallest rings is used. """ from rdkit.Chem.rdmolops import SanitizeMol from rdkit.Chem.rdchem import Mol mcopy = molecule.toRDKitMol(removeHs=True, returnMapping=False) SanitizeMol(mcopy) symmetryNumber = 1 # Get symmetry number for each ring in structure rings = molecule.getSmallestSetOfSmallestRings() for ring0 in rings: # Make copy of structure structure = molecule.copy(True) ring = [structure.atoms[molecule.atoms.index(atom)] for atom in ring0] # Figure out which atoms and bonds are aromatic and reassign appropriately: for i, atom1 in enumerate(ring0): for atom2 in ring0[i + 1:]: if molecule.hasBond(atom1, atom2): if mcopy.GetBondBetweenAtoms(i, i + 1) is not None: if str( mcopy.GetBondBetweenAtoms( i, i + 1).GetBondType()) == 'AROMATIC': bond = molecule.getBond(atom1, atom2) bond.applyAction( ['CHANGE_BOND', atom1, 'B', atom2]) atom1.atomType = atom2.atomType = rmgpy.molecule.atomTypes[ 'Cb'] else: pass # Remove bonds of ring from structure for i, atom1 in enumerate(ring): for atom2 in ring[i + 1:]: if structure.hasBond(atom1, atom2): structure.removeBond(atom1.edges[atom2]) structures = structure.split() groups = [] for struct in structures: for atom in ring: if struct.hasAtom(atom): struct.removeAtom(atom) groups.append(struct.split()) # Find equivalent functional groups on ring equivalentGroups = [] equivalentGroupCount = [] for group in groups: found = False for i, eqGroup in enumerate(equivalentGroups): if not found and len(group) == len(eqGroup): for g, eg in zip(group, eqGroup): if not g.isIsomorphic(eg): # The groups do not match break else: # The groups match found = True if found: # We've found a matching group, so increment its count equivalentGroupCount[i] += 1 break else: # No matching group found, so add it as a new group equivalentGroups.append(group) equivalentGroupCount.append(1) # Find equivalent bonds on ring equivalentBonds = [] for i, atom1 in enumerate(ring0): for atom2 in ring0[i + 1:]: if molecule.hasBond(atom1, atom2): bond = molecule.getBond(atom1, atom2) found = False for eqBond in equivalentBonds: if not found: if bond.equivalent(eqBond[0]): eqBond.append(group) found = True if not found: equivalentBonds.append([bond]) # Find maximum number of equivalent groups and bonds minEquivalentGroups = min(equivalentGroupCount) maxEquivalentGroups = max(equivalentGroupCount) minEquivalentBonds = None maxEquivalentBonds = 0 for bonds in equivalentBonds: N = len(bonds) if minEquivalentBonds is None or N < minEquivalentBonds: minEquivalentBonds = N if N > maxEquivalentBonds: maxEquivalentBonds = N if maxEquivalentGroups == maxEquivalentBonds == len(ring): symmetryNumber *= len(ring) * 2 else: symmetryNumber *= min(minEquivalentGroups, minEquivalentBonds) #print len(ring), minEquivalentGroups, maxEquivalentGroups, minEquivalentBonds, maxEquivalentBonds, symmetryNumber return symmetryNumber
def calculateCyclicSymmetryNumber(molecule): """ Get the symmetry number correction for cyclic regions of a molecule. For complicated fused rings the smallest set of smallest rings is used. """ from rdkit.Chem.rdmolops import SanitizeMol from rdkit.Chem.rdchem import Mol symmetryNumber = 1 rings = molecule.getSmallestSetOfSmallestRings() # Get symmetry number for each ring in structure for ring0 in rings: # Make another copy structure structure = molecule.copy(True) ring = [structure.atoms[molecule.atoms.index(atom)] for atom in ring0] # Remove bonds of ring from structure for i, atom1 in enumerate(ring): for atom2 in ring[i + 1:]: if structure.hasBond(atom1, atom2): structure.removeBond(atom1.edges[atom2]) structures = structure.split() groups = [] for struct in structures: for atom in ring: if struct.hasAtom(atom): struct.removeAtom(atom) groups.append(struct.split()) # Find equivalent functional groups on ring equivalentGroups = [] equivalentGroupCount = [] for group in groups: found = False for i, eqGroup in enumerate(equivalentGroups): if not found and len(group) == len(eqGroup): for g, eg in zip(group, eqGroup): if not g.isIsomorphic(eg): # The groups do not match break else: # The groups match found = True if found: # We've found a matching group, so increment its count equivalentGroupCount[i] += 1 break else: # No matching group found, so add it as a new group equivalentGroups.append(group) equivalentGroupCount.append(1) # Find equivalent bonds on ring equivalentBonds = [] for i, atom1 in enumerate(ring0): for atom2 in ring0[i + 1:]: if molecule.hasBond(atom1, atom2): bond = molecule.getBond(atom1, atom2) found = False for eqBond in equivalentBonds: if not found: if bond.equivalent(eqBond[0]): eqBond.append(group) found = True if not found: equivalentBonds.append([bond]) # Find maximum number of equivalent groups and bonds minEquivalentGroups = min(equivalentGroupCount) maxEquivalentGroups = max(equivalentGroupCount) minEquivalentBonds = None maxEquivalentBonds = 0 for bonds in equivalentBonds: N = len(bonds) if minEquivalentBonds is None or N < minEquivalentBonds: minEquivalentBonds = N if N > maxEquivalentBonds: maxEquivalentBonds = N if maxEquivalentGroups == maxEquivalentBonds == len(ring): symmetryNumber *= len(ring) * 2 else: symmetryNumber *= min(minEquivalentGroups, minEquivalentBonds) return symmetryNumber