def make_restraints(m, comp_id, mogul_dir, mogul_file_name_stub, pdb_out_file_name, mmcif_dict_name,
quartet_planes, quartet_hydrogen_planes, use_mmff,
match_atom_names_to_dict_flag,
comp_id_list_for_names_match,
dict_files_for_names_match):
# test here (or in calling functions) if m is sane (i.e. is an rdkit molecule)
if not isinstance(m, Chem.rdchem.Mol):
print 'ERROR:: not a molecule'
return False
n_attempts = 20 * m.GetNumAtoms() # default is 10 * number of atoms.
# pH-dependent protonation or deprotonation
#
do_hydrogen_atoms_shift = True
try:
compound_name = m.GetProp('_Name');
except KeyError:
# this happens all the time when we start from a SMILES, users don't need to see it.
# print 'caught key error in trying to get _Name in make_restraints() for m'
compound_name = '.'
except AttributeError as e:
# Do we need to see this? Perhaps make_restraints() needs to return a status.
# print 'AttributeError: problem with molecule in make_restraints()', e, ' on object:', m
return
m_H = m
if n_hydrogens(m) == 0:
m_H = AllChem.AddHs(m)
if do_hydrogen_atoms_shift:
# simple sane pH H-exchanges
sane_H_mol = pyrogen_boost.hydrogen_transformations(m_H)
# print >>file('sane_H.mol','w+'),Chem.MolToMolBlock(sane_H_mol)
else:
sane_H_mol = m_H
# This makes UFF types, which can fail sometimes.
conf_id = AllChem.EmbedMolecule(sane_H_mol, maxAttempts=n_attempts)
if use_mmff:
AllChem.MMFFOptimizeMolecule(sane_H_mol, confId=conf_id)
if False: # debugging output
ba = pyrogen_boost.mmff_bonds_and_angles(sane_H_mol) # uses _forcefield_ of the molecule
n_bonds = ba.bonds_size()
if n_bonds > 0:
for i_bond in range(n_bonds):
bond = ba.get_bond(i_bond)
print bond.get_idx_1(), bond.get_idx_2(), bond.get_type(), \
bond.get_resting_bond_length(), bond.get_sigma()
n_angles = ba.angles_size()
if n_angles > 0:
for i_angle in range(n_angles):
angle = ba.get_angle(i_angle)
print angle.get_idx_1(), angle.get_idx_2(), angle.get_idx_3(), \
angle.get_resting_angle(), angle.get_sigma()
else:
AllChem.UFFOptimizeMolecule(sane_H_mol, confId=conf_id)
# AllChem.UFFOptimizeMolecule(sane_H_mol)
atom_names = add_atom_names(sane_H_mol)
all_set = atom_types.set_atom_types(sane_H_mol) # has deloc bonds now, potentially
# debug sane_H_mol
if True:
molblock = Chem.MolToMolBlock(sane_H_mol)
print >> file("sane_H_mol.mol",'w'), molblock
if (all_set != True):
return False
else:
sane_H_mol.SetProp('comp_id', comp_id)
sane_H_mol.SetProp('name', compound_name)
sd_local = mogul_file_name_stub + ".sdf"
sdf_file_name = os.path.join(mogul_dir, mogul_file_name_stub + '-mogul.sdf')
mogul_ins_file_name = os.path.join(mogul_dir, mogul_file_name_stub + '-mogul.ins')
mogul_out_file_name = os.path.join(mogul_dir, mogul_file_name_stub + '-mogul.out')
Chem.AllChem.ComputeGasteigerCharges(sane_H_mol)
moguled_mol = pyrogen_boost.mogulify(sane_H_mol) # Nitro bond orders (and other things?)
if not os.path.isdir(mogul_dir):
checked_mkdir(mogul_dir)
if os.path.isdir(mogul_dir):
mb = Chem.MolToMolBlock(moguled_mol)
print >> file(sdf_file_name,'w'), mb
else:
mb = Chem.MolToMolBlock(moguled_mol)
print >> file(sdf_file_name,'w'), mb
bor = make_restraints_for_bond_orders(sane_H_mol)
# print out the set types:
print_atom_props = False
if print_atom_props:
print '--- Atom Props ---'
for atom in sane_H_mol.GetAtoms():
charge = atom.GetProp('_GasteigerCharge') # string?
name = atom.GetProp('name')
try:
atom_type = atom.GetProp('atom_type')
is_aromatic = atom.GetIsAromatic()
hybrid = atom.GetHybridization()
f_charge = float(charge)
if print_atom_props:
print " atom: %s %s type: %s arom: %s hybrid: %s charge: %6.3f" % (name, atom.GetSymbol(),
atom_type.ljust(4),
str(is_aromatic).ljust(5),
str(hybrid).rjust(3),
f_charge)
except KeyError:
print "miss", name, atom.GetSymbol(), charge
#
replace_with_mmff_b_a_restraints = False
if use_mmff:
replace_with_mmff_b_a_restraints = True
# execute_mogul() tests if mogul is executable
#
mogul_state = execute_mogul(sdf_file_name, mogul_ins_file_name, mogul_out_file_name)
if mogul_state:
# Here we need to think about matching to reference
# dictionary of amino acids (for standard atom names).
# That function takes a dictionary and a mmdb::Residue.
# How does that fit in here?
#
restraints = pysw.mogul_out_to_mmcif_dict_by_mol(mogul_out_file_name, comp_id,
compound_name, sane_H_mol, bor,
mmcif_dict_name, # not used
quartet_planes,
quartet_hydrogen_planes,
replace_with_mmff_b_a_restraints)
# match_atom_names_to_dict_flag, comp_id_list_for_names_match, dict_file_for_names_match
if match_atom_names_to_dict_flag:
restraints = atom_match_dictionary(restraints, sane_H_mol,
comp_id_list_for_names_match,
dict_files_for_names_match)
pysw.write_restraints(restraints, mmcif_dict_name)
pysw.regularize_and_write_pdb(sane_H_mol, restraints, comp_id, pdb_out_file_name)
else:
# mogul failed or was not in the path:
if run_mogul == False:
# ... but that's OK if we told pyrogen to run without mogul
# sane_H_mol:
# print >>file('debug_sane_H.mol','w+'),Chem.MolToMolBlock(sane_H_mol)
restraints = pysw.mmcif_dict_from_mol(comp_id, compound_name, sane_H_mol,
mmcif_dict_name,
quartet_planes, quartet_hydrogen_planes,
replace_with_mmff_b_a_restraints)
if restraints == None:
print "No restraints"
return True # hacked in value
if match_atom_names_to_dict_flag:
restraints = atom_match_dictionary(restraints, sane_H_mol,
comp_id_list_for_names_match,
dict_files_for_names_match)
pysw.write_restraints(restraints, mmcif_dict_name)
pysw.write_pdb_from_mol(sane_H_mol, comp_id, pdb_out_file_name)
else:
# ... but not if we wanted to use mogul.
# (We get here if there is a licence error for mogul)
exit(1)
return sane_H_mol
def make_restraints(m, comp_id, mogul_dir, mogul_file_name_stub, pdb_out_file_name, mmcif_dict_name,
quartet_planes, quartet_hydrogen_planes, use_mmff,
match_atom_names_to_dict_flag,
comp_id_list_for_names_match,
dict_files_for_names_match):
# test here (or in calling functions) if m is sane (i.e. is an rdkit molecule)
if not isinstance(m, Chem.rdchem.Mol):
print 'ERROR:: not a molecule'
return False
n_attempts = 20 * m.GetNumAtoms() # default is 10 * number of atoms.
# pH-dependent protonation or deprotonation
#
do_hydrogen_atoms_shift = True
try:
compound_name = m.GetProp('_Name');
except KeyError:
# this happens all the time when we start from a SMILES, users don't need to see it.
# print 'caught key error in trying to get _Name in make_restraints() for m'
compound_name = '.'
except AttributeError as e:
# Do we need to see this? Perhaps make_restraints() needs to return a status.
# print 'AttributeError: problem with molecule in make_restraints()', e, ' on object:', m
return
m_H = m
if n_hydrogens(m) == 0:
m_H = AllChem.AddHs(m)
if do_hydrogen_atoms_shift:
# simple sane pH H-exchanges
sane_H_mol = pyrogen_boost.hydrogen_transformations(m_H)
# print >>file('sane_H.mol','w+'),Chem.MolToMolBlock(sane_H_mol)
else:
sane_H_mol = m_H
# This makes UFF types, which can fail sometimes.
conf_id = AllChem.EmbedMolecule(sane_H_mol, maxAttempts=n_attempts)
if use_mmff:
AllChem.MMFFOptimizeMolecule(sane_H_mol, confId=conf_id)
if False: # debugging output
ba = pyrogen_boost.mmff_bonds_and_angles(sane_H_mol) # uses _forcefield_ of the molecule
n_bonds = ba.bonds_size()
if n_bonds > 0:
for i_bond in range(n_bonds):
bond = ba.get_bond(i_bond)
print bond.get_idx_1(), bond.get_idx_2(), bond.get_type(), \
bond.get_resting_bond_length(), bond.get_sigma()
n_angles = ba.angles_size()
if n_angles > 0:
for i_angle in range(n_angles):
angle = ba.get_angle(i_angle)
print angle.get_idx_1(), angle.get_idx_2(), angle.get_idx_3(), \
angle.get_resting_angle(), angle.get_sigma()
else:
AllChem.UFFOptimizeMolecule(sane_H_mol, confId=conf_id)
# AllChem.UFFOptimizeMolecule(sane_H_mol)
atom_names = add_atom_names(sane_H_mol)
all_set = atom_types.set_atom_types(sane_H_mol) # has deloc bonds now, potentially
# debug sane_H_mol
if True:
molblock = Chem.MolToMolBlock(sane_H_mol)
print >> file("sane_H_mol.mol",'w'), molblock
if (all_set != True):
return False
else:
sane_H_mol.SetProp('comp_id', comp_id)
sane_H_mol.SetProp('name', compound_name)
sd_local = mogul_file_name_stub + ".sdf"
sdf_file_name = os.path.join(mogul_dir, mogul_file_name_stub + '-mogul.sdf')
mogul_ins_file_name = os.path.join(mogul_dir, mogul_file_name_stub + '-mogul.ins')
mogul_out_file_name = os.path.join(mogul_dir, mogul_file_name_stub + '-mogul.out')
Chem.AllChem.ComputeGasteigerCharges(sane_H_mol)
moguled_mol = pyrogen_boost.mogulify(sane_H_mol) # Nitro bond orders (and other things?)
if not os.path.isdir(mogul_dir):
checked_mkdir(mogul_dir)
if os.path.isdir(mogul_dir):
mb = Chem.MolToMolBlock(moguled_mol)
print >> file(sdf_file_name,'w'), mb
else:
mb = Chem.MolToMolBlock(moguled_mol)
print >> file(sdf_file_name,'w'), mb
bor = make_restraints_for_bond_orders(sane_H_mol)
# print out the set types:
print_atom_props = False
if print_atom_props:
print '--- Atom Props ---'
for atom in sane_H_mol.GetAtoms():
charge = atom.GetProp('_GasteigerCharge') # string?
name = atom.GetProp('name')
try:
atom_type = atom.GetProp('atom_type')
is_aromatic = atom.GetIsAromatic()
hybrid = atom.GetHybridization()
f_charge = float(charge)
if print_atom_props:
print " atom: %s %s type: %s arom: %s hybrid: %s charge: %6.3f" % (name, atom.GetSymbol(),
atom_type.ljust(4),
str(is_aromatic).ljust(5),
str(hybrid).rjust(3),
f_charge)
except KeyError:
print "miss", name, atom.GetSymbol(), charge
#
replace_with_mmff_b_a_restraints = False
if use_mmff:
replace_with_mmff_b_a_restraints = True
# execute_mogul() tests if mogul is executable
#
mogul_state = execute_mogul(sdf_file_name, mogul_ins_file_name, mogul_out_file_name)
if mogul_state:
# Here we need to think about matching to reference
# dictionary of amino acids (for standard atom names).
# That function takes a dictionary and a mmdb::Residue.
# How does that fit in here?
#
restraints = pysw.mogul_out_to_mmcif_dict_by_mol(mogul_out_file_name, comp_id,
compound_name, sane_H_mol, bor,
mmcif_dict_name, # not used
quartet_planes,
quartet_hydrogen_planes,
replace_with_mmff_b_a_restraints)
# match_atom_names_to_dict_flag, comp_id_list_for_names_match, dict_file_for_names_match
if match_atom_names_to_dict_flag:
restraints = atom_match_dictionary(restraints, sane_H_mol,
comp_id_list_for_names_match,
dict_files_for_names_match)
pysw.write_restraints(restraints, mmcif_dict_name)
pysw.regularize_and_write_pdb(sane_H_mol, restraints, comp_id, pdb_out_file_name)
else:
# mogul failed or was not in the path:
if run_mogul == False:
# ... but that's OK if we told pyrogen to run without mogul
# sane_H_mol:
# print >>file('debug_sane_H.mol','w+'),Chem.MolToMolBlock(sane_H_mol)
restraints = pysw.mmcif_dict_from_mol(comp_id, compound_name, sane_H_mol,
mmcif_dict_name,
quartet_planes, quartet_hydrogen_planes,
replace_with_mmff_b_a_restraints)
if restraints == None:
print "No restraints"
return True # hacked in value
if match_atom_names_to_dict_flag:
restraints = atom_match_dictionary(restraints, sane_H_mol,
comp_id_list_for_names_match,
dict_files_for_names_match)
pysw.write_restraints(restraints, mmcif_dict_name)
pysw.write_pdb_from_mol(sane_H_mol, comp_id, pdb_out_file_name)
else:
# ... but not if we wanted to use mogul.
# (We get here if there is a licence error for mogul)
exit(1)
return sane_H_mol