def _generate_ffxmls():
import os
print(os.getcwd())
print("Parameterizing T4 inhibitors")
from perses.tests.testsystems import T4LysozymeInhibitorsTestSystem
from openmoltools import forcefield_generators
testsystem_t4 = T4LysozymeInhibitorsTestSystem()
smiles_list_t4 = testsystem_t4.molecules
oemols_t4 = [generate_molecule_from_smiles(smiles, idx=i) for i, smiles in enumerate(smiles_list_t4)]
ffxml_str_t4, failed_list = forcefield_generators.generateForceFieldFromMolecules(oemols_t4, ignoreFailures=True)
ffxml_out_t4 = open('/Users/grinawap/T4-inhibitors.xml','w')
ffxml_out_t4.write(ffxml_str_t4)
ffxml_out_t4.close()
if failed_list:
print("Failed some T4 inhibitors")
_print_failed_SMILES(failed_list)
print("Parameterizing kinase inhibitors")
from perses.tests.testsystems import KinaseInhibitorsTestSystem
testsystem_kinase = KinaseInhibitorsTestSystem()
smiles_list_kinase = testsystem_kinase.molecules
oemols_kinase = [generate_molecule_from_smiles(smiles, idx=i) for i, smiles in enumerate(smiles_list_kinase)]
ffxml_str_kinase, failed_kinase_list = forcefield_generators.generateForceFieldFromMolecules(oemols_kinase, ignoreFailures=True)
ffxml_out_kinase = open("/Users/grinawap/kinase-inhibitors.xml",'w')
ffxml_out_kinase.write(ffxml_str_kinase)
ffxml_out_t4.close()
def _generate_ffxmls():
import os
print(os.getcwd())
print("Parameterizing T4 inhibitors")
from perses.tests.testsystems import T4LysozymeInhibitorsTestSystem
from openmoltools import forcefield_generators
testsystem_t4 = T4LysozymeInhibitorsTestSystem()
smiles_list_t4 = testsystem_t4.molecules
oemols_t4 = [generate_molecule_from_smiles(smiles, idx=i) for i, smiles in enumerate(smiles_list_t4)]
ffxml_str_t4, failed_list = forcefield_generators.generateForceFieldFromMolecules(oemols_t4, ignoreFailures=True)
ffxml_out_t4 = open('/Users/grinawap/T4-inhibitors.xml','w')
ffxml_out_t4.write(ffxml_str_t4)
ffxml_out_t4.close()
if failed_list:
print("Failed some T4 inhibitors")
_print_failed_SMILES(failed_list)
print("Parameterizing kinase inhibitors")
from perses.tests.testsystems import KinaseInhibitorsTestSystem
testsystem_kinase = KinaseInhibitorsTestSystem()
smiles_list_kinase = testsystem_kinase.molecules
oemols_kinase = [generate_molecule_from_smiles(smiles, idx=i) for i, smiles in enumerate(smiles_list_kinase)]
ffxml_str_kinase, failed_kinase_list = forcefield_generators.generateForceFieldFromMolecules(oemols_kinase, ignoreFailures=True)
ffxml_out_kinase = open("/Users/grinawap/kinase-inhibitors.xml",'w')
ffxml_out_kinase.write(ffxml_str_kinase)
ffxml_out_t4.close()
def test_generate_ffxml_from_molecules():
"""
Test generation of single ffxml file from a list of molecules
"""
# Create a test set of molecules.
molecules = [createOEMolFromIUPAC(name) for name in IUPAC_molecule_names]
# Create an ffxml file.
from openmoltools.forcefield_generators import generateForceFieldFromMolecules
ffxml = generateForceFieldFromMolecules(molecules)
# Create a ForceField.
gaff_xml_filename = utils.get_data_filename("parameters/gaff.xml")
forcefield = ForceField(gaff_xml_filename)
try:
forcefield.loadFile(StringIO(ffxml))
except Exception as e:
msg = str(e)
msg += "ffxml contents:\n"
for (index, line) in enumerate(ffxml.split('\n')):
msg += 'line %8d : %s\n' % (index, line)
raise Exception(msg)
# Parameterize the molecules.
from openmoltools.forcefield_generators import generateTopologyFromOEMol
for molecule in molecules:
# Create topology from molecule.
topology = generateTopologyFromOEMol(molecule)
# Create system with forcefield.
system = forcefield.createSystem(topology)
# Check potential is finite.
positions = extractPositionsFromOEMOL(molecule)
check_potential_is_finite(system, positions)
def test_generate_ffxml_from_molecules():
"""
Test generation of single ffxml file from a list of molecules
"""
# Create a test set of molecules.
molecules = [createOEMolFromIUPAC(name) for name in IUPAC_molecule_names]
# Create an ffxml file.
from openmoltools.forcefield_generators import generateForceFieldFromMolecules
ffxml = generateForceFieldFromMolecules(molecules)
# Create a ForceField.
gaff_xml_filename = utils.get_data_filename("parameters/gaff.xml")
forcefield = ForceField(gaff_xml_filename)
try:
forcefield.loadFile(StringIO(ffxml))
except Exception as e:
msg = str(e)
msg += "ffxml contents:\n"
for (index, line) in enumerate(ffxml.split("\n")):
msg += "line %8d : %s\n" % (index, line)
raise Exception(msg)
# Parameterize the molecules.
from openmoltools.forcefield_generators import generateTopologyFromOEMol
for molecule in molecules:
# Create topology from molecule.
topology = generateTopologyFromOEMol(molecule)
# Create system with forcefield.
system = forcefield.createSystem(topology)
# Check potential is finite.
positions = extractPositionsFromOEMOL(molecule)
check_potential_is_finite(system, positions)
def write_xml(topology, file_name):
residues = [residue for residue in topology.residues()]
residue = residues[0]
molOE = generateOEMolFromTopologyResidue(residue,
geometry=False,
tripos_atom_names=True)
molOE.SetTitle('MOL')
ffxml = generateForceFieldFromMolecules([molOE])
f = open(file_name, 'w')
f.write(ffxml)
f.close()
def generate_ffxml(pdb_filename):
from simtk.openmm.app import PDBFile, Modeller
pdbfile = PDBFile(pdb_filename)
residues = [ residue for residue in pdbfile.topology.residues() ]
residue = residues[0]
from openmoltools.forcefield_generators import generateForceFieldFromMolecules, generateOEMolFromTopologyResidue
molecule = generateOEMolFromTopologyResidue(residue, geometry=False, tripos_atom_names=True)
molecule.SetTitle('MOL')
molecules = [molecule]
ffxml = generateForceFieldFromMolecules(molecules)
outfile = open('imatinib.xml', 'w')
outfile.write(ffxml)
outfile.close()
def generate_ffxml(pdb_filename):
from simtk.openmm.app import PDBFile, Modeller
pdbfile = PDBFile(pdb_filename)
residues = [residue for residue in pdbfile.topology.residues()]
residue = residues[0]
from openmoltools.forcefield_generators import generateForceFieldFromMolecules, generateOEMolFromTopologyResidue
molecule = generateOEMolFromTopologyResidue(residue,
geometry=False,
tripos_atom_names=True)
molecule.SetTitle('MOL')
molecules = [molecule]
ffxml = generateForceFieldFromMolecules(molecules)
outfile = open('imatinib.xml', 'w')
outfile.write(ffxml)
outfile.close()
def generate_forcefield(molecule_file, outfile):
ifs = oechem.oemolistream()
ifs.open(molecule_file)
# get the list of molecules
mol_list = [normalize_molecule(oechem.OEMol(mol)) for mol in ifs.GetOEMols()]
# TODO: HORRIBLE HACK ; WILL DIE AT > 999 RESIDUES!
for idx, mol in enumerate(mol_list):
mol.SetTitle("%03d" % idx)
ffxml = forcefield_generators.generateForceFieldFromMolecules(mol_list, normalize=False)
with open(outfile, 'w') as output_file:
output_file.write(ffxml)
def generate_forcefield(molecule_file, outfile):
ifs = oechem.oemolistream()
ifs.open(molecule_file)
# get the list of molecules
mol_list = [
normalize_molecule(oechem.OEMol(mol)) for mol in ifs.GetOEMols()
]
# TODO: HORRIBLE HACK ; WILL DIE AT > 999 RESIDUES!
for idx, mol in enumerate(mol_list):
mol.SetTitle("%03d" % idx)
ffxml = forcefield_generators.generateForceFieldFromMolecules(
mol_list, normalize=False)
with open(outfile, 'w') as output_file:
output_file.write(ffxml)
#!/usr/bin/env python
"""
Generate ffxml files for ligands in a multi-mole mol2.
"""
mol2_filename = 'Imatinib-epik-charged.mol2'
ffxml_filename = 'Imatinib-epik-charged.ffxml'
# Read mol2 file containing protonation states and extract canonical isomeric SMILES from this.
print("Reading molecules")
from openeye import oechem
ifs = oechem.oemolistream(mol2_filename)
mol = oechem.OEMol()
molecules = list()
while oechem.OEReadMolecule(ifs, mol):
molecules.append(oechem.OEMol(mol))
print("Generating forcefield parameters...")
from openmoltools.forcefield_generators import generateForceFieldFromMolecules
ffxml = generateForceFieldFromMolecules(molecules,
ignoreFailures=False,
generateUniqueNames=True)
print("Writing forcefield to '%s'..." % ffxml_filename)
outfile = open(ffxml_filename, 'w')
outfile.write(ffxml)
outfile.close()
#!/usr/bin/env python
"""
Generate ffxml files for ligands in a multi-mole mol2.
"""
import openeye
import openmoltools
mol2_filename = 'Imatinib-epik-charged.mol2'
ffxml_filename = 'Imatinib-epik-charged.ffxml'
# Read mol2 file containing protonation states and extract canonical isomeric SMILES from this.
print("Reading molecules")
from openeye import oechem
ifs = oechem.oemolistream(mol2_filename)
mol = oechem.OEMol()
molecules = list()
while oechem.OEReadMolecule(ifs, mol):
molecules.append(oechem.OEMol(mol))
print("Generating forcefield parameters...")
from openmoltools.forcefield_generators import generateForceFieldFromMolecules
ffxml = generateForceFieldFromMolecules(molecules, ignoreFailures=False, generateUniqueNames=True)
print("Writing forcefield to '%s'..." % ffxml_filename)
outfile = open(ffxml_filename, 'w')
outfile.write(ffxml)
outfile.close()
def generate_solvated_hybrid_test_topology(current_mol_name="naphthalene",
proposed_mol_name="benzene",
current_mol_smiles=None,
proposed_mol_smiles=None,
vacuum=False,
render_atom_mapping=False):
"""
This function will generate a topology proposal, old positions, and new positions with a geometry proposal (either vacuum or solvated) given a set of input iupacs or smiles.
The function will (by default) read the iupac names first. If they are set to None, then it will attempt to read a set of current and new smiles.
An atom mapping pdf will be generated if specified.
Arguments
----------
current_mol_name : str, optional
name of the first molecule
proposed_mol_name : str, optional
name of the second molecule
current_mol_smiles : str (default None)
current mol smiles
proposed_mol_smiles : str (default None)
proposed mol smiles
vacuum: bool (default False)
whether to render a vacuum or solvated topology_proposal
render_atom_mapping : bool (default False)
whether to render the atom map of the current_mol_name and proposed_mol_name
Returns
-------
topology_proposal : perses.rjmc.topology_proposal
The topology proposal representing the transformation
current_positions : np.array, unit-bearing
The positions of the initial system
new_positions : np.array, unit-bearing
The positions of the new system
"""
import simtk.openmm.app as app
from openmoltools import forcefield_generators
from openeye import oechem
from openmoltools.openeye import iupac_to_oemol, generate_conformers, smiles_to_oemol
from openmoltools import forcefield_generators
import perses.utils.openeye as openeye
from perses.utils.data import get_data_filename
from perses.rjmc.topology_proposal import TopologyProposal, SystemGenerator, SmallMoleculeSetProposalEngine
import simtk.unit as unit
from perses.rjmc.geometry import FFAllAngleGeometryEngine
if current_mol_name != None and proposed_mol_name != None:
try:
old_oemol, new_oemol = iupac_to_oemol(
current_mol_name), iupac_to_oemol(proposed_mol_name)
old_smiles = oechem.OECreateSmiString(
old_oemol,
oechem.OESMILESFlag_DEFAULT | oechem.OESMILESFlag_Hydrogens)
new_smiles = oechem.OECreateSmiString(
new_oemol,
oechem.OESMILESFlag_DEFAULT | oechem.OESMILESFlag_Hydrogens)
except:
raise Exception(
f"either {current_mol_name} or {proposed_mol_name} is not compatible with 'iupac_to_oemol' function!"
)
elif current_mol_smiles != None and proposed_mol_smiles != None:
try:
old_oemol, new_oemol = smiles_to_oemol(
current_mol_smiles), smiles_to_oemol(proposed_mol_smiles)
old_smiles = oechem.OECreateSmiString(
old_oemol,
oechem.OESMILESFlag_DEFAULT | oechem.OESMILESFlag_Hydrogens)
new_smiles = oechem.OECreateSmiString(
new_oemol,
oechem.OESMILESFlag_DEFAULT | oechem.OESMILESFlag_Hydrogens)
except:
raise Exception(f"the variables are not compatible")
else:
raise Exception(
f"either current_mol_name and proposed_mol_name must be specified as iupacs OR current_mol_smiles and proposed_mol_smiles must be specified as smiles strings."
)
old_oemol, old_system, old_positions, old_topology = openeye.createSystemFromSMILES(
old_smiles, title="MOL")
#correct the old positions
old_positions = openeye.extractPositionsFromOEMol(old_oemol)
old_positions = old_positions.in_units_of(unit.nanometers)
new_oemol, new_system, new_positions, new_topology = openeye.createSystemFromSMILES(
new_smiles, title="NEW")
ffxml = forcefield_generators.generateForceFieldFromMolecules(
[old_oemol, new_oemol])
old_oemol.SetTitle('MOL')
new_oemol.SetTitle('MOL')
old_topology = forcefield_generators.generateTopologyFromOEMol(old_oemol)
new_topology = forcefield_generators.generateTopologyFromOEMol(new_oemol)
if not vacuum:
nonbonded_method = app.PME
barostat = openmm.MonteCarloBarostat(1.0 * unit.atmosphere,
300.0 * unit.kelvin, 50)
else:
nonbonded_method = app.NoCutoff
barostat = None
gaff_xml_filename = get_data_filename("data/gaff.xml")
system_generator = SystemGenerator(
[gaff_xml_filename, 'amber99sbildn.xml', 'tip3p.xml'],
barostat=barostat,
forcefield_kwargs={
'removeCMMotion': False,
'nonbondedMethod': nonbonded_method,
'constraints': app.HBonds,
'hydrogenMass': 4.0 * unit.amu
})
system_generator._forcefield.loadFile(StringIO(ffxml))
proposal_engine = SmallMoleculeSetProposalEngine([old_smiles, new_smiles],
system_generator,
residue_name='MOL')
geometry_engine = FFAllAngleGeometryEngine(metadata=None,
use_sterics=False,
n_bond_divisions=1000,
n_angle_divisions=180,
n_torsion_divisions=360,
verbose=True,
storage=None,
bond_softening_constant=1.0,
angle_softening_constant=1.0,
neglect_angles=False)
if not vacuum:
#now to solvate
modeller = app.Modeller(old_topology, old_positions)
hs = [
atom for atom in modeller.topology.atoms()
if atom.element.symbol in ['H']
and atom.residue.name not in ['MOL', 'OLD', 'NEW']
]
modeller.delete(hs)
modeller.addHydrogens(forcefield=system_generator._forcefield)
modeller.addSolvent(system_generator._forcefield,
model='tip3p',
padding=9.0 * unit.angstroms)
solvated_topology = modeller.getTopology()
solvated_positions = modeller.getPositions()
solvated_positions = unit.quantity.Quantity(value=np.array([
list(atom_pos) for atom_pos in
solvated_positions.value_in_unit_system(unit.md_unit_system)
]),
unit=unit.nanometers)
solvated_system = system_generator.build_system(solvated_topology)
#now to create proposal
top_proposal = proposal_engine.propose(
current_system=solvated_system,
current_topology=solvated_topology,
current_mol=old_oemol,
proposed_mol=new_oemol)
new_positions, _ = geometry_engine.propose(top_proposal,
solvated_positions, beta)
if render_atom_mapping:
from perses.utils.smallmolecules import render_atom_mapping
print(
f"new_to_old: {proposal_engine.non_offset_new_to_old_atom_map}"
)
render_atom_mapping(f"{old_smiles}to{new_smiles}.png", old_oemol,
new_oemol,
proposal_engine.non_offset_new_to_old_atom_map)
return top_proposal, solvated_positions, new_positions
else:
vacuum_system = system_generator.build_system(old_topology)
top_proposal = proposal_engine.propose(current_system=vacuum_system,
current_topology=old_topology,
current_mol=old_oemol,
proposed_mol=new_oemol)
new_positions, _ = geometry_engine.propose(top_proposal, old_positions,
beta)
if render_atom_mapping:
from perses.utils.smallmolecules import render_atom_mapping
print(f"new_to_old: {top_proposal._new_to_old_atom_map}")
render_atom_mapping(f"{old_smiles}to{new_smiles}.png", old_oemol,
new_oemol, top_proposal._new_to_old_atom_map)
return top_proposal, old_positions, new_positions
def __init__(self,
protein_pdb_filename,
ligand_file,
old_ligand_index,
new_ligand_index,
forcefield_files,
pressure=1.0 * unit.atmosphere,
temperature=300.0 * unit.kelvin,
solvent_padding=9.0 * unit.angstroms):
"""
Initialize a NonequilibriumFEPSetup object
Parameters
----------
protein_pdb_filename : str
The name of the protein pdb file
ligand_file : str
the name of the ligand file (any openeye supported format)
ligand_smiles : list of two str
The SMILES strings representing the two ligands
forcefield_files : list of str
The list of ffxml files that contain the forcefields that will be used
pressure : Quantity, units of pressure
Pressure to use in the barostat
temperature : Quantity, units of temperature
Temperature to use for the Langevin integrator
solvent_padding : Quantity, units of length
The amount of padding to use when adding solvent
"""
self._protein_pdb_filename = protein_pdb_filename
self._pressure = pressure
self._temperature = temperature
self._barostat_period = 50
self._padding = solvent_padding
self._ligand_file = ligand_file
self._old_ligand_index = old_ligand_index
self._new_ligand_index = new_ligand_index
self._old_ligand_oemol = self.load_sdf(self._ligand_file,
index=self._old_ligand_index)
self._new_ligand_oemol = self.load_sdf(self._ligand_file,
index=self._new_ligand_index)
self._old_ligand_positions = extractPositionsFromOEMOL(
self._old_ligand_oemol)
ffxml = forcefield_generators.generateForceFieldFromMolecules(
[self._old_ligand_oemol, self._new_ligand_oemol])
self._old_ligand_oemol.SetTitle("MOL")
self._new_ligand_oemol.SetTitle("MOL")
self._new_ligand_smiles = oechem.OECreateSmiString(
self._new_ligand_oemol,
oechem.OESMILESFlag_DEFAULT | oechem.OESMILESFlag_Hydrogens)
#self._old_ligand_smiles = '[H]c1c(c(c(c(c1N([H])c2nc3c(c(n2)OC([H])([H])C4(C(C(C(C(C4([H])[H])([H])[H])([H])[H])([H])[H])([H])[H])[H])nc(n3[H])[H])[H])[H])S(=O)(=O)C([H])([H])[H])[H]'
self._old_ligand_smiles = oechem.OECreateSmiString(
self._old_ligand_oemol,
oechem.OESMILESFlag_DEFAULT | oechem.OESMILESFlag_Hydrogens)
print(self._new_ligand_smiles)
print(self._old_ligand_smiles)
self._old_ligand_topology = forcefield_generators.generateTopologyFromOEMol(
self._old_ligand_oemol)
self._old_ligand_md_topology = md.Topology.from_openmm(
self._old_ligand_topology)
self._new_ligand_topology = forcefield_generators.generateTopologyFromOEMol(
self._new_ligand_oemol)
self._new_liands_md_topology = md.Topology.from_openmm(
self._new_ligand_topology)
protein_pdbfile = open(self._protein_pdb_filename, 'r')
pdb_file = app.PDBFile(protein_pdbfile)
protein_pdbfile.close()
self._protein_topology_old = pdb_file.topology
self._protein_md_topology_old = md.Topology.from_openmm(
self._protein_topology_old)
self._protein_positions_old = pdb_file.positions
self._forcefield = app.ForceField(*forcefield_files)
self._forcefield.loadFile(StringIO(ffxml))
print("Generated forcefield")
self._complex_md_topology_old = self._protein_md_topology_old.join(
self._old_ligand_md_topology)
self._complex_topology_old = self._complex_md_topology_old.to_openmm()
n_atoms_complex_old = self._complex_topology_old.getNumAtoms()
n_atoms_protein_old = self._protein_topology_old.getNumAtoms()
self._complex_positions_old = unit.Quantity(np.zeros(
[n_atoms_complex_old, 3]),
unit=unit.nanometers)
self._complex_positions_old[:
n_atoms_protein_old, :] = self._protein_positions_old
self._complex_positions_old[
n_atoms_protein_old:, :] = self._old_ligand_positions
if pressure is not None:
barostat = openmm.MonteCarloBarostat(self._pressure,
self._temperature,
self._barostat_period)
self._system_generator = SystemGenerator(
forcefield_files,
barostat=barostat,
forcefield_kwargs={'nonbondedMethod': app.PME})
else:
self._system_generator = SystemGenerator(forcefield_files)
#self._complex_proposal_engine = TwoMoleculeSetProposalEngine(self._old_ligand_smiles, self._new_ligand_smiles, self._system_generator, residue_name="MOL")
self._complex_proposal_engine = TwoMoleculeSetProposalEngine(
self._old_ligand_oemol,
self._new_ligand_oemol,
self._system_generator,
residue_name="MOL")
self._geometry_engine = FFAllAngleGeometryEngine()
self._complex_topology_old_solvated, self._complex_positions_old_solvated, self._complex_system_old_solvated = self._solvate_system(
self._complex_topology_old, self._complex_positions_old)
self._complex_md_topology_old_solvated = md.Topology.from_openmm(
self._complex_topology_old_solvated)
print(self._complex_proposal_engine._smiles_list)
beta = 1.0 / (kB * temperature)
self._complex_topology_proposal = self._complex_proposal_engine.propose(
self._complex_system_old_solvated,
self._complex_topology_old_solvated)
self._complex_positions_new_solvated, _ = self._geometry_engine.propose(
self._complex_topology_proposal,
self._complex_positions_old_solvated, beta)
#now generate the equivalent objects for the solvent phase. First, generate the ligand-only topologies and atom map
self._solvent_topology_proposal, self._old_solvent_positions = self._generate_ligand_only_topologies(
self._complex_positions_old_solvated,
self._complex_positions_new_solvated)
self._new_solvent_positions, _ = self._geometry_engine.propose(
self._solvent_topology_proposal, self._old_solvent_positions, beta)