def RingSubSet(ifs, ofs, exo):
for mol in ifs.GetOEGraphMols():
submol = oechem.OEGraphMol()
adjustHcount = True
if exo:
isinring = oechem.OEAtomIsInRing()
isexo = oechem.OEIsNonRingAtomDoubleBondedToRing()
includeexo = oechem.OEOrAtom(isinring, isexo)
oechem.OESubsetMol(submol, mol, includeexo, adjustHcount)
else:
oechem.OESubsetMol(submol, mol, oechem.OEAtomIsInRing(),
adjustHcount)
submol.SetTitle(mol.GetTitle() + "_rings")
if submol.NumAtoms() != 0:
oechem.OEWriteMolecule(ofs, submol)
def get_binding_site(structure, ligand):
"""
"""
partlist = oechem.OEIntArray(structure.GetMaxAtomIdx())
# get all contacts between ligand and assembly
contacts = oechem.OEGetNearestNbrs(structure, ligand, 5.0)
binding_site_atom_idxs = set(contact.GetBgn().GetIdx()
for contact in contacts)
# create the partition map
for atom in structure.GetAtoms():
if atom.GetIdx() in binding_site_atom_idxs:
partlist[atom.GetIdx()] = 1
entitypred = oechem.OEPartPredAtom(partlist)
# select the binding site atoms
entitypred.SelectPart(1)
# create a new molecule for the entity
binding_site = oechem.OEGraphMol()
oechem.OESubsetMol(binding_site, structure, entitypred, False, False)
return binding_site
def AtomPathLength(ifs, ofs, itf, atm1, atm2):
for mol in ifs.GetOEGraphMols():
oechem.OETriposAtomNames(mol)
a1 = None
a2 = None
for atm in mol.GetAtoms():
if atm.GetName() == atm1:
a1 = atm
if atm.GetName() == atm2:
a2 = atm
if a1 is not None and a2 is not None:
break
if a1 is None or a2 is None:
oechem.OEThrow.Warning(
"Failed to find atoms %s and %s in molecule" % (atm1, atm2))
continue
pathlen = oechem.OEGetPathLength(a1, a2)
if itf.GetBool("-verbose") or not itf.HasString("-o"):
print("Path length: %s in %s" %
(pathlen, oechem.OEMolToSmiles(mol)))
spath = oechem.OEShortestPath(a1, a2)
spathmol = oechem.OEGraphMol()
adjustHCount = True
oechem.OESubsetMol(spathmol, mol, oechem.OEIsAtomMember(spath),
adjustHCount)
spathsmiles = oechem.OEMolToSmiles(spathmol)
if itf.HasString("-o"):
oechem.OEWriteMolecule(ofs, spathmol)
elif itf.GetBool("-verbose"):
print(spathsmiles)
def atom_bond_set_to_mol(frag, oemol, adjust_hcount=False, RGroup=True):
from openeye import oechem
import warnings
fragatompred = oechem.OEIsAtomMember(frag.GetAtoms())
fragbondpred = oechem.OEIsBondMember(frag.GetBonds())
fragment_oemol = oechem.OEMol()
adjustHCount = adjust_hcount
oechem.OESubsetMol(fragment_oemol, oemol, fragatompred, fragbondpred,
adjustHCount, RGroup)
oechem.OEAddExplicitHydrogens(fragment_oemol)
# sanity check that all atoms are bonded
for atom in fragment_oemol.GetAtoms():
if not list(atom.GetBonds()):
warnings.warn(
"Yikes!!! An atom that is not bonded to any other atom in the fragment. "
"You probably ran into a bug. Please report the input molecule to the issue tracker"
)
# Perceive stereo and check that defined stereo did not change
oechem.OEPerceiveChiral(fragment_oemol)
oechem.OE3DToAtomStereo(fragment_oemol)
oechem.OE3DToBondStereo(fragment_oemol)
return fragment_oemol
def split_molecule_components(
molecule: oechem.OEGraphMol) -> List[oechem.OEGraphMol]:
"""
Split an OpenEye Molecule into its bonded components.
Parameters
----------
molecule: oechem.OEGraphMol
An OpenEye molecule holding multiple components.
Returns
-------
: list of oechem.OEGraphMol
A list of OpenEye molecules holding the split components.
"""
# determine bonded components
number_of_components, part_list = oechem.OEDetermineComponents(molecule)
predicate = oechem.OEPartPredAtom(part_list)
# get bonded components
components = []
for i in range(number_of_components):
predicate.SelectPart(i + 1)
component = oechem.OEGraphMol()
oechem.OESubsetMol(component, molecule, predicate)
components.append(component)
return components
def frag_to_smiles(frags, mol):
"""
Convert fragments (AtomBondSet) to smiles string
Parameters
----------
frags
mol
Returns
-------
smiles: list of smiles strings
"""
smiles = {}
for frag in frags:
fragatompred = oechem.OEIsAtomMember(frag.GetAtoms())
fragbondpred = oechem.OEIsBondMember(frag.GetBonds())
fragment = oechem.OEGraphMol()
adjustHCount = True
oechem.OESubsetMol(fragment, mol, fragatompred, fragbondpred,
adjustHCount)
s = oechem.OECreateIsoSmiString(fragment)
if s not in smiles:
smiles[s] = []
smiles[s].append(frag)
return smiles
def get_scaffold(molecule, adjustHcount=False):
"""
Takes an openeye.oechem.oemol and returns
an openeye.oechem.oemol of the scaffold
The scaffold is a molecule where all the atoms that are not in rings, and are not linkers between rings.
double bonded atoms exo to a ring are included as ring atoms
This function has been completely taken from openeye's extractscaffold.py script
https://docs.eyesopen.com/toolkits/python/oechemtk/oechem_examples/oechem_example_extractscaffold.html#section-example-oechem-extractscaffold
Parameters
----------
mol : openeye.oechem.oemol
entire molecule to get the scaffold of
adjustHcount : bool, default=False
add/remove hydrogens to satisfy valence of scaffold
Returns
-------
openeye.oechem.oemol
scaffold oemol of the input mol. New oemol.
"""
def TraverseForRing(visited, atom):
visited.add(atom.GetIdx())
for nbor in atom.GetAtoms():
if nbor.GetIdx() not in visited:
if nbor.IsInRing():
return True
if TraverseForRing(visited, nbor):
return True
return False
def DepthFirstSearchForRing(root, nbor):
visited = set()
visited.add(root.GetIdx())
return TraverseForRing(visited, nbor)
class IsInScaffold(oechem.OEUnaryAtomPred):
def __call__(self, atom):
if atom.IsInRing():
return True
count = 0
for nbor in atom.GetAtoms():
if DepthFirstSearchForRing(atom, nbor):
count += 1
return count > 1
dst = oechem.OEMol()
pred = IsInScaffold()
oechem.OESubsetMol(dst, molecule, pred, adjustHcount)
return dst
def search(self, mol0, mol1):
mol0 = mol0._struc
mol1 = mol1._struc
p0 = mol0.CreateCopy()
p1 = mol1.CreateCopy()
#set atom int type.
for mol in (
p0,
p1,
):
for atom in mol.GetAtoms():
if (atom.IsHydrogen()):
atom.SetIntType(1)
else:
atom.SetIntType(2)
#suppress hydrogens before mcs search
oechem.OESuppressHydrogens(p0)
oechem.OESuppressHydrogens(p1)
if (self._is_approximate):
mcss = oechem.OEMCSSearch(p1, self._atom_expr, self._bond_expr,
oechem.OEMCSType_Approximate)
else:
mcss = oechem.OEMCSSearch(p1, self._atom_expr, self._bond_expr)
#set minimum atom of the mcs
mcss.SetMinAtoms(1)
#set the function to evalue the mcs search
mcss.SetMCSFunc(oechem.OEMCSMaxAtomsCompleteCycles(1.5))
# There could be multiple matches. We select the one with the maximum number of atoms.
# If there are more than 1 matches with the same maximum number of atoms, we arbitrarily select the first one.
mcs_mol = None
max_num = 0
#do the mcs search
for match in mcss.Match(p0, True):
num_atom = 0
mcs_tmp = oechem.OEMol()
oechem.OESubsetMol(mcs_tmp, match, True)
oechem.OEFindRingAtomsAndBonds(mcs_tmp)
for atom in mcs_tmp.GetAtoms():
if (not atom.IsHydrogen()):
num_atom += 1
if (num_atom > max_num):
max_num = num_atom
mcs_mol = mcs_tmp
atom_match0 = []
atom_match1 = []
for matchpair in match.GetAtoms():
atom_match0.append(matchpair.target.GetIdx() + 1)
atom_match1.append(matchpair.pattern.GetIdx() + 1)
#dump search result to kbase
if (mcs_mol):
mol0 = struc.OeStruc(mol0)
mol1 = struc.OeStruc(mol1)
mcs_mol = struc.OeStruc(mcs_mol)
return self.deposit_to_kbase(mol0.id(), mol1.id(), atom_match0,
atom_match1)
def strip_water_ions(in_system):
"""
This function remove waters and ions molecules
from the input system
Parameters:
----------
in_system : oechem.OEMol
The bio-molecular system to clean
opt: python dictionary
The system option
Output:
-------
clean_system : oechem.OEMol
The cleaned system
"""
# Copy the input system
system = in_system.CreateCopy()
# Create a bit vector mask
bv = oechem.OEBitVector(system.GetMaxAtomIdx())
bv.NegateBits()
# Create a Hierarchical View of the protein system
hv = oechem.OEHierView(
system,
oechem.OEAssumption_BondedResidue + oechem.OEAssumption_ResPerceived)
# Looping over the system residues
for chain in hv.GetChains():
for frag in chain.GetFragments():
for hres in frag.GetResidues():
res = hres.GetOEResidue()
# Check if a residue is a mono atomic ion
natoms = 0
for at in hres.GetAtoms():
natoms += 1
# Set the atom bit mask off
if oechem.OEGetResidueIndex(
res) == oechem.OEResidueIndex_HOH or natoms == 1:
# Set Bit mask
atms = hres.GetAtoms()
for at in atms:
bv.SetBitOff(at.GetIdx())
# Extract the system without waters or ions
pred = oechem.OEAtomIdxSelected(bv)
clean_system = oechem.OEMol()
oechem.OESubsetMol(clean_system, system, pred)
return clean_system
def SmartsPathLength(ifs, ofs, itf, ss1, ss2):
for mol in ifs.GetOEGraphMols():
oechem.OEPrepareSearch(mol, ss1)
oechem.OEPrepareSearch(mol, ss2)
if not (ss1.SingleMatch(mol) and ss2.SingleMatch(mol)):
oechem.OEThrow.Warning(
"Unable to find SMARTS matches in %s, skipping" %
mol.GetTitle())
continue
unique = True
allminlen = sys.maxsize
for match1 in ss1.Match(mol, unique):
for match2 in ss2.Match(mol, unique):
minlen = sys.maxsize
for atom1 in match1.GetTargetAtoms():
for atom2 in match2.GetTargetAtoms():
pathlen = oechem.OEGetPathLength(atom1, atom2)
if minlen > pathlen:
minlen = pathlen
atompairs = []
atompairs.append([atom1, atom2])
elif minlen == pathlen:
atompairs.append([atom1, atom2])
if minlen < allminlen:
allminlen = minlen
allatompairs = atompairs[:]
elif minlen == allminlen:
allatompairs += atompairs[:]
if itf.GetBool("-verbose") or not itf.HasString("-o"):
print("Shortest path length: %s in %s" %
(allminlen, oechem.OEMolToSmiles(mol)))
spathlist = set()
for satom1, satom2, in allatompairs:
spath = oechem.OEShortestPath(satom1, satom2)
spathmol = oechem.OEGraphMol()
oechem.OESubsetMol(spathmol, mol, oechem.OEIsAtomMember(spath))
spathsmiles = oechem.OEMolToSmiles(spathmol)
if spathsmiles in spathlist:
continue
spathlist.add(spathsmiles)
if itf.HasString("-o"):
oechem.OEWriteMolecule(ofs, spathmol)
elif itf.GetBool("-verbose"):
print(spathsmiles)
return
def get_smarts(prefix, atom_idxs):
"""Get the SMARTS corresponding to a list of atom indices"""
offmol = Molecule.from_file(prefix + '.mol2')
fix_carboxylate_bond_orders(offmol)
if prefix in prefix2pmd_struct:
pmd_struct = prefix2pmd_struct[prefix]
else:
pmd_struct = ParmEd.load_file(prefix + '.prmtop')
prefix2pmd_struct[prefix] = pmd_struct
oemol = offmol.to_openeye()
residues_of_interest = set()
atom_indices_of_interest = set()
#for atom_idx in atom_idxs:
#residues_of_interest.add(pmd_struct.atoms[atom_idx].residue.idx)
#atom_indices_of_interest.add(atom_idx)
#for neighbor in oemol.GetAtom(atom_idx).GetAtoms():
# atom_indices_of_interest.add(neighbor.GetIdx())
for oeatom, pmd_atom in zip(oemol.GetAtoms(), pmd_struct.atoms):
# Delete all non-residue-of-interest atoms
#if (pmd_atom.residue.idx in residues_of_interest):
# atom_indices_of_interest.add(pmd_atom.idx)
# Assign tags to atoms of interest
if (oeatom.GetIdx() in atom_idxs):
atom_idx = oeatom.GetIdx()
map_index = atom_idxs.index(atom_idx) + 1
oeatom.SetMapIdx(map_index)
atom_indices_of_interest.add(atom_idx)
for neighbor in oeatom.GetAtoms():
atom_indices_of_interest.add(neighbor.GetIdx())
# Make a "Subset" molecule, so that we don't get weird charges
# around where we cleave the residues
subsetmol = OEChem.OEGraphMol()
oepred = OEChem.PyAtomPredicate(
lambda x: x.GetIdx() in atom_indices_of_interest)
OEChem.OESubsetMol(subsetmol, oemol, oepred)
smiles_options = (OEChem.OESMILESFlag_Canonical
| OEChem.OESMILESFlag_Isotopes
| OEChem.OESMILESFlag_RGroups)
# Add the atom and bond stereo flags
smiles_options |= OEChem.OESMILESFlag_AtomStereo | OEChem.OESMILESFlag_BondStereo
# Add the hydrogen flag
smiles_options |= OEChem.OESMILESFlag_Hydrogens
smiles_options |= OEChem.OESMILESFlag_AtomMaps
smiles = OEChem.OECreateSmiString(subsetmol, smiles_options)
return smiles
def GetFragmentCombinations(mol, fraglist, frag_number):
fragments = []
fragcombs = GetFragmentAtomBondSetCombinations(mol, fraglist, frag_number)
for f in fragcombs:
fragatompred = oechem.OEIsAtomMember(f.GetAtoms())
fragbondpred = oechem.OEIsBondMember(f.GetBonds())
fragment = oechem.OEGraphMol()
adjustHCount = True
oechem.OESubsetMol(fragment, mol, fragatompred, fragbondpred,
adjustHCount)
fragments.append(fragment)
return fragments
def BackBone(ifs, ofs):
adjustHCount = True
mol = oechem.OEGraphMol()
backboneMol = oechem.OEGraphMol()
while oechem.OEReadMolecule(ifs, mol):
if not oechem.OEHasResidues(mol):
oechem.OEPerceiveResidues(mol, oechem.OEPreserveResInfo_All)
aiter = mol.GetAtoms(oechem.OEIsBackboneAtom())
member = oechem.OEIsAtomMember(aiter)
oechem.OESubsetMol(backboneMol, mol, member, adjustHCount)
oechem.OEWriteMolecule(ofs, backboneMol)
def GetFragments(mol, minbonds, maxbonds):
from openeye import oegraphsim
frags = []
fptype = oegraphsim.OEGetFPType("Tree,ver=2.0.0,size=4096,bonds=%d-%d,atype=AtmNum,btype=Order"
% (minbonds, maxbonds))
for abset in oegraphsim.OEGetFPCoverage(mol, fptype, True):
fragatompred = oechem.OEIsAtomMember(abset.GetAtoms())
frag = oechem.OEGraphMol()
adjustHCount = True
oechem.OESubsetMol(frag, mol, fragatompred, adjustHCount)
oechem.OEFindRingAtomsAndBonds(frag)
frags.append(oechem.OEGraphMol(frag))
return frags
def get_atom_map(tagged_smiles, molecule=None):
"""
Returns a dictionary that maps tag on SMILES to atom index in molecule.
Parameters
----------
tagged_smiles: str
index-tagged explicit hydrogen SMILES string
molecule: OEMol
molecule to generate map for. If None, a new OEMol will be generated from the tagged SMILES, the map will map to
this molecule and it will be returned.
Returns
-------
atom_map: dict
a dictionary that maps tag to atom index {tag:idx}
molecule: OEMol
If a molecule was not provided, the generated molecule will be returned.
"""
if molecule is None:
molecule = openeye.smiles_to_oemol(tagged_smiles)
ss = oechem.OESubSearch(tagged_smiles)
oechem.OEPrepareSearch(molecule, ss)
ss.SetMaxMatches(1)
atom_map = {}
t1 = time.time()
matches = [m for m in ss.Match(molecule)]
t2 = time.time()
seconds = t2 - t1
logger().info("Substructure search took {} seconds".format(seconds))
if not matches:
logger().info("MCSS failed for {}, smiles: {}".format(
molecule.GetTitle(), tagged_smiles))
return False
for match in matches:
for ma in match.GetAtoms():
atom_map[ma.pattern.GetMapIdx()] = ma.target.GetIdx()
# sanity check
mol = oechem.OEGraphMol()
oechem.OESubsetMol(mol, match, True)
logger().info("Match SMILES: {}".format(oechem.OEMolToSmiles(mol)))
if molecule is None:
return molecule, atom_map
return atom_map
def SubSetRes(ifs, ofs, chainid, resname, resnum):
adjustHCount = True
mol = oechem.OEGraphMol()
while oechem.OEReadMolecule(ifs, mol):
if not oechem.OEHasResidues(mol):
oechem.OEPerceiveResidues(mol, oechem.OEPreserveResInfo_All)
hv = oechem.OEHierView(mol)
res = hv.GetResidue(chainid, resname, resnum)
if res.GetOEResidue().GetName() is None:
oechem.OEThrow.Fatal("Failed to find residue")
atomiter = res.GetAtoms()
member = oechem.OEIsAtomMember(atomiter)
resmol = oechem.OEGraphMol()
oechem.OESubsetMol(resmol, mol, member, adjustHCount)
if chainid == " ":
resmol.SetTitle("%s %d" % (resname, resnum))
else:
resmol.SetTitle("%s %s %d" % (resname, chainid, resnum))
oechem.OEWriteMolecule(ofs, resmol)
def frag_to_smiles(frags, mol):
"""
Convert fragments (AtomBondSet) to canonical isomeric SMILES string
Parameters
----------
frags: list
mol: OEMol
OESMILESFlag: str
Either 'ISOMERIC' or 'DEFAULT'. This flag determines which OE function to use to generate SMILES string
Returns
-------
smiles: dict of smiles to frag
"""
smiles = {}
for frag in frags:
fragatompred = oechem.OEIsAtomMember(frag.GetAtoms())
fragbondpred = oechem.OEIsBondMember(frag.GetBonds())
#fragment = oechem.OEGraphMol()
fragment = oechem.OEMol()
adjustHCount = True
oechem.OESubsetMol(fragment, mol, fragatompred, fragbondpred, adjustHCount)
oechem.OEPerceiveChiral(fragment)
# sanity check that all atoms are bonded
for atom in fragment.GetAtoms():
if not list(atom.GetBonds()):
raise Warning("Yikes!!! An atom that is not bonded to any other atom in the fragment. "
"You probably ran into a bug. Please report the input molecule to the issue tracker")
#s = oechem.OEMolToSmiles(fragment)
#s2 = fragmenter.utils.create_mapped_smiles(fragment, tagged=False, explicit_hydrogen=False)
s = mol_to_smiles(fragment, mapped=False, explicit_hydrogen=True, isomeric=True)
if s not in smiles:
smiles[s] = []
smiles[s].append(frag)
return smiles
def main(argv=[__name__]):
if len(argv) != 3:
oechem.OEThrow.Usage("%s <infile> <outfile>" % argv[0])
ifs = oechem.oemolistream()
if not ifs.open(argv[1]):
oechem.OEThrow.Fatal("Unable to open %s for reading" % argv[1])
ofs = oechem.oemolostream()
if not ofs.open(argv[2]):
oechem.OEThrow.Fatal("Unable to open %s for writing" % argv[2])
for mol in ifs.GetOEGraphMols():
numparts, partlist = oechem.OEDetermineComponents(mol)
pred = oechem.OEPartPredAtom(partlist)
for i in range(1, numparts + 1):
pred.SelectPart(i)
partmol = oechem.OEGraphMol()
oechem.OESubsetMol(partmol, mol, pred)
oechem.OEWriteMolecule(ofs, partmol)
def around(dist, ls):
"""
This function select atom not far than the threshold distance from
the current selection. The threshold distance is in Angstrom
selection can be:
mask = '5.0 around ligand'
"""
# at = system.GetAtom(oechem.OEHasAtomIdx(idx))
# Atom set selection
atom_set_around = set()
# Create a OE bit vector mask for each atoms
bv_around = oechem.OEBitVector(system.GetMaxAtomIdx())
# Set the mask atom
for at in system.GetAtoms():
if at.GetIdx() in ls:
bv_around.SetBitOn(at.GetIdx())
# Predicate
pred = oechem.OEAtomIdxSelected(bv_around)
# Create the system molecule based on the atom mask
molecules = oechem.OEMol()
oechem.OESubsetMol(molecules, system, pred)
# Create the Nearest neighbours
nn = oechem.OENearestNbrs(system, float(dist))
for nbrs in nn.GetNbrs(molecules):
for atom in oechem.OEGetResidueAtoms(nbrs.GetBgn()):
if atom.GetIdx() in ls:
continue
atom_set_around.add(atom.GetIdx())
return atom_set_around
def main(argv=[__name__]):
itf = oechem.OEInterface(InterfaceData, argv)
exo_dbl_bonds = itf.GetBool("-exo")
ifs = oechem.oemolistream()
if not ifs.open(itf.GetString("-i")):
oechem.OEThrow.Fatal("Unable to open %s for reading" % itf.GetString("-i"))
ofs = oechem.oemolostream()
if not ofs.open(itf.GetString("-o")):
oechem.OEThrow.Fatal("Unable to open %s for writing" % itf.GetString("-o"))
for src in ifs.GetOEMols():
dst = oechem.OEMol()
pred = IsInScaffold()
if exo_dbl_bonds:
pred = oechem.OEOrAtom(pred, oechem.OEIsNonRingAtomDoubleBondedToRing())
adjustHcount = True
oechem.OESubsetMol(dst, src, pred, adjustHcount)
if dst.IsValid():
oechem.OEWriteMolecule(ofs, dst)
#!/usr/bin/env python
# (C) 2017 OpenEye Scientific Software Inc. All rights reserved.
#
# TERMS FOR USE OF SAMPLE CODE The software below ("Sample Code") is
# provided to current licensees or subscribers of OpenEye products or
# SaaS offerings (each a "Customer").
# Customer is hereby permitted to use, copy, and modify the Sample Code,
# subject to these terms. OpenEye claims no rights to Customer's
# modifications. Modification of Sample Code is at Customer's sole and
# exclusive risk. Sample Code may require Customer to have a then
# current license or subscription to the applicable OpenEye offering.
# THE SAMPLE CODE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
# EXPRESS OR IMPLIED. OPENEYE DISCLAIMS ALL WARRANTIES, INCLUDING, BUT
# NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
# PARTICULAR PURPOSE AND NONINFRINGEMENT. In no event shall OpenEye be
# liable for any damages or liability in connection with the Sample Code
# or its use.
# @ <SNIPPET>
from openeye import oechem
mol = oechem.OEGraphMol()
oechem.OESmilesToMol(mol, "c1ccccc1O")
frag = oechem.OEGraphMol()
oechem.OESubsetMol(frag, mol, oechem.OEIsCarbon())
mol.SetData("just_carbon", frag)
justCarbon = mol.GetData("just_carbon")
# @ </SNIPPET>
# SaaS offerings (each a "Customer").
# Customer is hereby permitted to use, copy, and modify the Sample Code,
# subject to these terms. OpenEye claims no rights to Customer's
# modifications. Modification of Sample Code is at Customer's sole and
# exclusive risk. Sample Code may require Customer to have a then
# current license or subscription to the applicable OpenEye offering.
# THE SAMPLE CODE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
# EXPRESS OR IMPLIED. OPENEYE DISCLAIMS ALL WARRANTIES, INCLUDING, BUT
# NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
# PARTICULAR PURPOSE AND NONINFRINGEMENT. In no event shall OpenEye be
# liable for any damages or liability in connection with the Sample Code
# or its use.
# @ <SNIPPET>
from __future__ import print_function
from openeye import oechem
from openeye import oemedchem
mol = oechem.OEGraphMol()
oechem.OESmilesToMol(mol, "COc1ccc(cc1)CC(=O)N")
for frag in oemedchem.OEGetRingChainFragments(mol):
fragatompred = oechem.OEIsAtomMember(frag.GetAtoms())
fragbondpred = oechem.OEIsBondMember(frag.GetBonds())
fragment = oechem.OEGraphMol()
adjustHCount = True
oechem.OESubsetMol(fragment, mol, fragatompred, fragbondpred, adjustHCount)
print(oechem.OEMolToSmiles(fragment))
# @ </SNIPPET>
def get_atom_map(molecule, mapped_smiles, strict=True):
"""
Map tag in mapped SMILES to atom idx using a substructure search
A substructure search finds chemically equivalent matches so if atoms are symmetrical, they can flip. The mapped
SMILES used for the pattern is first used to generate a molecule with the map indices, the order is canonicalized
and then it is used for the substructure search pattern. This ensures that symmetrical do not flip but there is no
guarantee that it won't happen.
Parameters
----------
molecule: oechem.OEMOl
Must have explicit hydrogen
mapped_smiles: str
explicit hydrogen SMILES with map indices on every atom
Returns
-------
atom_map: dict
{map_idx:atom_idx}
"""
# check that smiles has explicit hydrogen and map indices
mapped_mol = oechem.OEMol()
oechem.OESmilesToMol(mapped_mol, mapped_smiles)
if not has_atom_map(mapped_mol):
raise ValueError(
"Mapped SMILES must have map indices for all atoms and hydrogens")
# Check molecule for explicit hydrogen
if not has_explicit_hydrogen(molecule) and strict:
raise ValueError("Molecule must have explicit hydrogens")
# canonical order mapped mol to ensure atom map is always generated in the same order
canonical_order_atoms(mapped_mol)
aopts = oechem.OEExprOpts_DefaultAtoms
bopts = oechem.OEExprOpts_DefaultBonds
ss = oechem.OESubSearch(mapped_mol, aopts, bopts)
oechem.OEPrepareSearch(molecule, ss)
ss.SetMaxMatches(1)
atom_map = {}
matches = [m for m in ss.Match(molecule)]
if not matches:
raise RuntimeError("MCSS failed for {}, smiles: {}".format(
oechem.OEMolToSmiles(molecule), mapped_smiles))
for match in matches:
for ma in match.GetAtoms():
atom_map[ma.pattern.GetMapIdx()] = ma.target.GetIdx()
# sanity check
mol = oechem.OEGraphMol()
oechem.OESubsetMol(mol, match, True)
matched_smiles = mol_to_smiles(mol,
isomeric=False,
explicit_hydrogen=False,
mapped=False)
molcopy = oechem.OEMol(molecule)
smiles = mol_to_smiles(molcopy,
isomeric=False,
explicit_hydrogen=False,
mapped=False)
pattern_smiles = mol_to_smiles(mapped_mol,
isomeric=False,
explicit_hydrogen=False,
mapped=False)
if not matched_smiles == smiles == pattern_smiles:
raise RuntimeError(
"Matched molecule, input molecule and mapped SMILES are not the same "
)
return atom_map
def generate_restricted_conformers(receptor, refmol, mol, core_smarts=None):
"""
Generate and select a conformer of the specified molecule using the reference molecule
Parameters
----------
receptor : openeye.oechem.OEGraphMol
Receptor (already prepped for docking) for identifying optimal pose
refmol : openeye.oechem.OEGraphMol
Reference molecule which shares some part in common with the proposed molecule
mol : openeye.oechem.OEGraphMol
Molecule whose conformers are to be enumerated
core_smarts : str, optional, default=None
If core_smarts is specified, substructure will be extracted using SMARTS.
"""
from openeye import oechem, oeomega
# DEBUG: For benzotriazoles, truncate refmol
core_smarts = 'c1ccc(NC(=O)[C,N]n2nnc3ccccc32)cc1' # prospective
core_smarts = 'NC(=O)[C,N]n2nnc3ccccc32' # retrospective
# Get core fragment
if core_smarts:
# Truncate refmol to SMARTS if specified
#print(f'Trunctating using SMARTS {refmol_smarts}')
ss = oechem.OESubSearch(core_smarts)
oechem.OEPrepareSearch(refmol, ss)
for match in ss.Match(refmol):
core_fragment = oechem.OEGraphMol()
oechem.OESubsetMol(core_fragment, match)
break
#print(f'refmol has {refmol.NumAtoms()} atoms')
else:
core_fragment = GetCoreFragment(refmol, [mol])
oechem.OESuppressHydrogens(core_fragment)
#print(f' Core fragment has {core_fragment.NumAtoms()} heavy atoms')
MIN_CORE_ATOMS = 6
if core_fragment.NumAtoms() < MIN_CORE_ATOMS:
return None
# Create an Omega instance
#omegaOpts = oeomega.OEOmegaOptions()
omegaOpts = oeomega.OEOmegaOptions(oeomega.OEOmegaSampling_Dense)
# Set the fixed reference molecule
omegaFixOpts = oeomega.OEConfFixOptions()
omegaFixOpts.SetFixMaxMatch(10) # allow multiple MCSS matches
omegaFixOpts.SetFixDeleteH(True) # only use heavy atoms
omegaFixOpts.SetFixMol(core_fragment)
#omegaFixOpts.SetFixSmarts(smarts)
omegaFixOpts.SetFixRMS(0.5)
atomexpr = oechem.OEExprOpts_Aromaticity | oechem.OEExprOpts_Hybridization
bondexpr = oechem.OEExprOpts_BondOrder | oechem.OEExprOpts_Aromaticity
omegaFixOpts.SetAtomExpr(atomexpr)
omegaFixOpts.SetBondExpr(bondexpr)
omegaOpts.SetConfFixOptions(omegaFixOpts)
molBuilderOpts = oeomega.OEMolBuilderOptions()
molBuilderOpts.SetStrictAtomTypes(False) # don't give up if MMFF types are not found
omegaOpts.SetMolBuilderOptions(molBuilderOpts)
omegaOpts.SetWarts(False) # expand molecule title
omegaOpts.SetStrictStereo(False) # set strict stereochemistry
omegaOpts.SetIncludeInput(False) # don't include input
omegaOpts.SetMaxConfs(1000) # generate lots of conformers
#omegaOpts.SetEnergyWindow(10.0) # allow high energies
omega = oeomega.OEOmega(omegaOpts)
from openeye import oequacpac
if not oequacpac.OEGetReasonableProtomer(mol):
print('No reasonable protomer found')
return None
mol = oechem.OEMol(mol) # multi-conformer molecule
ret_code = omega.Build(mol)
if (mol.GetDimension() != 3) or (ret_code != oeomega.OEOmegaReturnCode_Success):
print(f'Omega failure: {mol.GetDimension()} and {oeomega.OEGetOmegaError(ret_code)}')
return None
# Extract poses
class Pose(object):
def __init__(self, conformer):
self.conformer = conformer
self.clash_score = None
self.docking_score = None
self.overlap_score = None
poses = [ Pose(conf) for conf in mol.GetConfs() ]
# Score clashes
bump_check = BumpCheck(receptor)
for pose in poses:
pose.clash_score = bump_check.count(pose.conformer)
# Score docking poses
from openeye import oedocking
score = oedocking.OEScore(oedocking.OEScoreType_Chemgauss4)
score.Initialize(receptor)
for pose in poses:
pose.docking_score = score.ScoreLigand(pose.conformer)
# Compute overlap scores
from openeye import oeshape
overlap_prep = oeshape.OEOverlapPrep()
overlap_prep.Prep(refmol)
shapeFunc = oeshape.OEExactShapeFunc()
shapeFunc.SetupRef(refmol)
oeshape_result = oeshape.OEOverlapResults()
for pose in poses:
tmpmol = oechem.OEGraphMol(pose.conformer)
overlap_prep.Prep(tmpmol)
shapeFunc.Overlap(tmpmol, oeshape_result)
pose.overlap_score = oeshape_result.GetRefTversky()
# Filter poses based on top 10% of overlap
poses = sorted(poses, key= lambda pose : pose.overlap_score)
poses = poses[int(0.9*len(poses)):]
# Select the best docking score
import numpy as np
poses = sorted(poses, key=lambda pose : pose.docking_score)
pose = poses[0]
mol.SetActive(pose.conformer)
oechem.OESetSDData(mol, 'clash_score', str(pose.clash_score))
oechem.OESetSDData(mol, 'docking_score', str(pose.docking_score))
oechem.OESetSDData(mol, 'overlap_score', str(pose.overlap_score))
# Convert to single-conformer molecule
mol = oechem.OEGraphMol(mol)
return mol
def oesolvate(solute,
density=1.0,
padding_distance=10.0,
distance_between_atoms=2.5,
solvents='tip3p',
molar_fractions='1.0',
geometry='box',
close_solvent=True,
salt='[Na+], [Cl-]',
salt_concentration=0.0,
neutralize_solute=True,
verbose=False,
return_components=False,
**kargs):
"""
This function solvates the passed solute in a cubic box or a sphere by using Packmol. Packmol
creates an initial point for molecular dynamics simulations by packing molecule in defined regions
of space. For additional info:
http://www.ime.unicamp.br/~martinez/packmol/home.shtml
The geometry volume is estimated by the using the padding parameter and the solute size.
The number of solvent molecules is calculated by using the specified density and volume.
Solvent molecules are specified as comma separated smiles strings. The molar fractions
of each solvent molecule are specified in a similar fashion. By default if the solute is
charged counter ions are added to neutralize it
Parameters:
-----------
solute: OEMol molecule
The solute to solvate
density: float
The solution density in g/ml
padding_distance: float
The largest dimension of the solute (along the x, y, or z axis) is determined (in A),
and a cubic box of size (largest dimension)+2*padding is used
distance_between_atoms: float
The minimum distance between atoms in A
solvents: python string
A comma separated smiles string or keywords for the solvent molecules.
Special water models can be selected by using the keywords:
tip3p for TIP3P water model geometry
molar_fractions: python string
A comma separated molar fraction string of the solvent molecules
close_solvent: boolean
If True solvent molecules will be placed very close to the solute
salt: python string
A comma separated string of the dissociated salt in solution
salt_concentration: float
Salt concentration in millimolar
neutralize_solute: boolean
If True counter-ions will be added to the solution to neutralize the solute
verbose: Bool
If True verbose mode is enabled
return_components: Bool
If True the added solvent molecules are also returned as OEMol
Return:
-------
oe_mol: OEMol
The solvated system. If the selected geometry is a box a SD tag with
name 'box_vector' is attached the output molecule containing
the system box vectors.
oe_mol_components: OEMol
If the return_components flag is True the added solvent molecules are
returned as an additional OEMol
"""
def BoundingBox(molecule):
"""
This function calculates the Bounding Box of the passed
molecule
molecule: OEMol
return: bb (numpy array)
the calculated bounding box is returned as numpy array:
[(xmin,ymin,zmin), (xmax,ymax,zmax)]
"""
coords = [v for k, v in molecule.GetCoords().items()]
np_coords = np.array(coords)
min_coord = np_coords.min(axis=0)
max_coord = np_coords.max(axis=0)
bb = np.array([min_coord, max_coord])
return bb
if shutil.which("packmol") is None:
raise (IOError("Packmol executable not found"))
# Extract solvent smiles strings and mole fractions
solvents = [sm.strip() for sm in solvents.split(',')]
fractions = [float(mf) for mf in molar_fractions.split(',')]
# If the smiles string and mole fractions lists have different lengths raise an error
if len(solvents) != len(fractions):
raise ValueError(
"Selected solvent number and selected molar fraction number mismatch: {} vs {}"
.format(len(solvents), len(fractions)))
# Remove smiles string with 0.0 mole fraction
solvent_smiles = [
solvents[i] for i, v in enumerate(fractions) if fractions[i]
]
mol_fractions = [mf for mf in fractions if mf]
# Mole fractions are non-negative numbers
if any([v < 0.0 for v in mol_fractions]):
raise ValueError("Error: Mole fractions are non-negative real numbers")
# Mole fractions must sum up to 1.0
if abs(sum(mol_fractions) - 1.0) > 0.001:
oechem.OEThrow.Error("Error: Mole fractions do not sum up to 1.0")
if geometry not in ['box', 'sphere']:
raise ValueError(
"Error geometry: the supported geometries are box and sphere not {}"
.format(geometry))
# Set Units
density = density * unit.grams / unit.milliliter
padding_distance = padding_distance * unit.angstrom
salt_concentration = salt_concentration * unit.millimolar
# Calculate the Solute Bounding Box
BB_solute = BoundingBox(solute)
# Estimate of the box cube length
box_edge = 2.0 * padding_distance + np.max(BB_solute[1] -
BB_solute[0]) * unit.angstrom
if geometry == 'box':
# Box Volume
Volume = box_edge**3
if geometry == 'sphere':
Volume = (4.0 / 3.0) * 3.14159265 * (0.5 * box_edge)**3
# Omega engine is used to generate conformations
omegaOpts = oeomega.OEOmegaOptions()
omegaOpts.SetMaxConfs(1)
omegaOpts.SetStrictStereo(False)
omega = oeomega.OEOmega(omegaOpts)
# Create a string code to identify the solute residues. The code ID used is based
# on the residue number id, the residue name and the chain id:
# id+resname+chainID
hv_solute = oechem.OEHierView(
solute,
oechem.OEAssumption_BondedResidue + oechem.OEAssumption_ResPerceived)
solute_resid_list = []
for chain in hv_solute.GetChains():
for frag in chain.GetFragments():
for hres in frag.GetResidues():
oe_res = hres.GetOEResidue()
solute_resid_list.append(
str(oe_res.GetResidueNumber()) + oe_res.GetName() +
chain.GetChainID())
# Solvent component list_names
solvent_resid_dic_names = dict()
# Neutralize solute
ion_sum_wgt_n_ions = 0.0 * unit.grams / unit.mole
if neutralize_solute:
# Container for the counter-ions
oe_ions = []
# Container for the ion smiles strings
ions_smiles = []
solute_formal_charge = 0
for at in solute.GetAtoms():
solute_formal_charge += at.GetFormalCharge()
if solute_formal_charge > 0:
ions_smiles.append("[Cl-]")
elif solute_formal_charge < 0:
ions_smiles.append("[Na+]")
else:
pass
# Total number of counter-ions to neutralize the solute
n_ions = abs(solute_formal_charge)
# print("Counter ions to add = {} of {}".format(n_ions, ions_smiles[0]))
# Ions
if n_ions >= 1:
for sm in ions_smiles:
mol = oechem.OEMol()
if not oechem.OESmilesToMol(mol, sm):
raise ValueError(
"Error counter ions: SMILES string parsing fails for the string: {}"
.format(sm))
# Generate conformer
if not omega(mol):
raise ValueError(
"Error counter ions: Conformer generation fails for the molecule with "
"smiles string: {}".format(sm))
oe_ions.append(mol)
if sm == '[Na+]':
solvent_resid_dic_names[' NA'] = mol
else:
solvent_resid_dic_names[' CL'] = mol
ion_sum_wgt = 0.0 * unit.grams / unit.mole
for ion in oe_ions:
# Molecular weight
ion_sum_wgt += oechem.OECalculateMolecularWeight(
ion) * unit.grams / unit.mole
ion_sum_wgt_n_ions = ion_sum_wgt * n_ions
# Create ions .pdb files
ions_smiles_pdbs = []
for i in range(0, len(ions_smiles)):
pdb_name = os.path.basename(tempfile.mktemp(suffix='.pdb'))
pdb_name = ions_smiles[i] + '_' + pdb_name
ions_smiles_pdbs.append(pdb_name)
for i in range(0, len(ions_smiles)):
ofs = oechem.oemolostream(ions_smiles_pdbs[i])
oechem.OEWriteConstMolecule(ofs, oe_ions[i])
# Add salts to the solution
# Solvent smiles string parsing
char_set = string.ascii_uppercase
salt_sum_wgt_n_salt = 0.0 * unit.grams / unit.mole
if salt_concentration > 0.0 * unit.millimolar:
salt_smiles = [sm.strip() for sm in salt.split(',')]
# Container list of oemol salt molecules generated by using smiles strings
oe_salt = []
for sm in salt_smiles:
mol_salt = oechem.OEMol()
if not oechem.OESmilesToMol(mol_salt, sm):
raise ValueError(
"Error salt: SMILES string parsing fails for the string: {}"
.format(sm))
# Generate conformer
if not omega(mol_salt):
raise ValueError(
"Error salt: Conformer generation fails for the "
"molecule with smiles string: {}".format(sm))
# Unique 3 code letter are set as solvent residue names
solv_id = ''.join(random.sample(char_set * 3, 3))
# Try to recognize the residue name
oechem.OEPerceiveResidues(mol_salt)
for atmol in mol_salt.GetAtoms():
res = oechem.OEAtomGetResidue(atmol)
if res.GetName() == 'UNL':
res.SetName(solv_id)
oechem.OEAtomSetResidue(atmol, res)
if solv_id not in solvent_resid_dic_names:
solvent_resid_dic_names[solv_id] = mol_salt
else:
if res.GetName() not in solvent_resid_dic_names:
solvent_resid_dic_names[res.GetName()] = mol_salt
break
oe_salt.append(mol_salt)
n_salt = int(
round(unit.AVOGADRO_CONSTANT_NA * salt_concentration *
Volume.in_units_of(unit.liter)))
# for i in range(0, len(salt_smiles)):
# print("Number of molecules for the salt component {} = {}".format(salt_smiles[i], n_salt))
salt_sum_wgt = 0.0 * unit.grams / unit.mole
for salt in oe_salt:
# Molecular weight
salt_sum_wgt += oechem.OECalculateMolecularWeight(
salt) * unit.grams / unit.mole
salt_sum_wgt_n_salt = salt_sum_wgt * n_salt
# Create salt .pdb files
if n_salt >= 1:
salt_pdbs = []
for i in range(0, len(salt_smiles)):
pdb_name = os.path.basename(tempfile.mktemp(suffix='.pdb'))
# pdb_name = salt_smiles[i] + '_' + pdb_name
salt_pdbs.append(pdb_name)
for i in range(0, len(salt_smiles)):
ofs = oechem.oemolostream(salt_pdbs[i])
oechem.OEWriteConstMolecule(ofs, oe_salt[i])
# Container list of oemol solvent molecules generated by using smiles strings
oe_solvents = []
for sm in solvent_smiles:
if sm == 'tip3p':
tip3p_fn = os.path.join(PACKAGE_DIR, 'oeommtools', 'data',
'tip3p.pdb')
ifs = oechem.oemolistream(tip3p_fn)
mol_sol = oechem.OEMol()
if not oechem.OEReadMolecule(ifs, mol_sol):
raise IOError(
"It was not possible to read the tip3p molecule file")
else:
mol_sol = oechem.OEMol()
if not oechem.OESmilesToMol(mol_sol, sm):
raise ValueError(
"Error solvent: SMILES string parsing fails for the string: {}"
.format(sm))
# Generate conformer
if not omega(mol_sol):
raise ValueError(
"Error solvent: Conformer generation fails for "
"the molecule with smiles string: {}".format(sm))
# Unique 3 code letter are set as solvent residue names
solv_id = ''.join(random.sample(char_set * 3, 3))
# Try to recognize the residue name
oechem.OEPerceiveResidues(mol_sol)
for atmol in mol_sol.GetAtoms():
res = oechem.OEAtomGetResidue(atmol)
if res.GetName() == 'UNL':
res.SetName(solv_id)
oechem.OEAtomSetResidue(atmol, res)
if solv_id not in solvent_resid_dic_names:
solvent_resid_dic_names[solv_id] = mol_sol
else:
if res.GetName() not in solvent_resid_dic_names:
solvent_resid_dic_names[res.GetName()] = mol_sol
break
oe_solvents.append(mol_sol)
# Sum of the solvent molecular weights
solvent_sum_wgt_frac = 0.0 * unit.grams / unit.mole
for idx in range(0, len(oe_solvents)):
# Molecular weight
wgt = oechem.OECalculateMolecularWeight(
oe_solvents[idx]) * unit.grams / unit.mole
solvent_sum_wgt_frac += wgt * mol_fractions[idx]
# Solute molecular weight
solute_wgt = oechem.OECalculateMolecularWeight(
solute) * unit.gram / unit.mole
# Estimate of the number of each molecular species present in the solution accordingly
# to their molar fraction fi:
#
# ni = fi*(density*volume*NA - wgt_solute - sum_k(wgt_salt_k*nk) - wgt_ion*n_ion)/sum_j(wgt_nj * fj)
#
# where ni is the number of molecule of specie i, density the mixture density, volume the
# mixture volume, wgt_solute the molecular weight of the solute, wgt_salt_k the molecular
# weight of the salt component k, nk the number of molecule of salt component k, wgt_ion
# the counter ion molecular weight, n_ions the number of counter ions and wgt_nj the molecular
# weight of the molecule specie j with molar fraction fj
div = (unit.AVOGADRO_CONSTANT_NA * density * Volume -
(solute_wgt + salt_sum_wgt_n_salt +
ion_sum_wgt_n_ions)) / solvent_sum_wgt_frac
# Solvent number of monomers
n_monomers = [int(round(mf * div)) for mf in mol_fractions]
if not all([nm > 0 for nm in n_monomers]):
raise ValueError(
"Error negative number of solvent components: the density could be too low"
)
# for i in range(0, len(solvent_smiles)):
# print("Number of molecules for the component {} = {}".format(solvent_smiles[i], n_monomers[i]))
# Packmol Configuration file setting
if close_solvent:
header_template = """\n# Mixture\ntolerance {}\nfiletype pdb\noutput {}\nadd_amber_ter\navoid_overlap no"""
else:
header_template = """\n# Mixture\ntolerance {}\nfiletype pdb\noutput {}\nadd_amber_ter\navoid_overlap yes"""
# Templates strings
solute_template = """\n\n# Solute\nstructure {}\nnumber 1\nfixed 0. 0. 0. 0. 0. 0.\nresnumbers 1\nend structure"""
if geometry == 'box':
solvent_template = """\nstructure {}\nnumber {}\ninside box {:0.3f} {:0.3f} {:0.3f} {:0.3f} {:0.3f} {:0.3f}\
\nchain !\nresnumbers 3\nend structure"""
if geometry == 'sphere':
solvent_template = """\nstructure {}\nnumber {}\ninside sphere {:0.3f} {:0.3f} {:0.3f} {:0.3f}\
\nchain !\nresnumbers 3\nend structure"""
# Create solvents .pdb files
solvent_pdbs = []
for i in range(0, len(solvent_smiles)):
pdb_name = os.path.basename(tempfile.mktemp(suffix='.pdb'))
solvent_pdbs.append(pdb_name)
for i in range(0, len(solvent_smiles)):
ofs = oechem.oemolostream(solvent_pdbs[i])
oechem.OEWriteConstMolecule(ofs, oe_solvents[i])
solute_pdb = 'solute' + '_' + os.path.basename(
tempfile.mktemp(suffix='.pdb'))
ofs = oechem.oemolostream(solute_pdb)
if solute.GetMaxConfIdx() > 1:
raise ValueError("Solutes with multiple conformers are not supported")
else:
oechem.OEWriteConstMolecule(ofs, solute)
# Write Packmol header section
mixture_pdb = 'mixture' + '_' + os.path.basename(
tempfile.mktemp(suffix='.pdb'))
body = header_template.format(distance_between_atoms, mixture_pdb)
# Write Packmol configuration file solute section
body += solute_template.format(solute_pdb)
# The solute is centered inside the box
xc = (BB_solute[0][0] + BB_solute[1][0]) / 2.
yc = (BB_solute[0][1] + BB_solute[1][1]) / 2.
zc = (BB_solute[0][2] + BB_solute[1][2]) / 2.
# Correct for periodic box conditions to avoid
# steric clashes at the box edges
pbc_correction = 1.0 * unit.angstrom
xmin = xc - ((box_edge - pbc_correction) / 2.) / unit.angstrom
xmax = xc + ((box_edge - pbc_correction) / 2.) / unit.angstrom
ymin = yc - ((box_edge - pbc_correction) / 2.) / unit.angstrom
ymax = yc + ((box_edge - pbc_correction) / 2.) / unit.angstrom
zmin = zc - ((box_edge - pbc_correction) / 2.) / unit.angstrom
zmax = zc + ((box_edge - pbc_correction) / 2.) / unit.angstrom
# Packmol setting for the solvent section
body += '\n\n# Solvent'
for i in range(0, len(solvent_smiles)):
if geometry == 'box':
body += solvent_template.format(solvent_pdbs[i], n_monomers[i],
xmin, ymin, zmin, xmax, ymax, zmax)
if geometry == 'sphere':
body += solvent_template.format(solvent_pdbs[i], n_monomers[i], xc,
yc, zc,
0.5 * box_edge / unit.angstrom)
# Packmol setting for the salt section
if salt_concentration > 0.0 * unit.millimolar and n_salt >= 1:
body += '\n\n# Salt'
for i in range(0, len(salt_smiles)):
if geometry == 'box':
body += solvent_template.format(salt_pdbs[i],
int(round(n_salt)), xmin, ymin,
zmin, xmax, ymax, zmax)
if geometry == 'sphere':
body += solvent_template.format(salt_pdbs[i],
int(round(n_salt)), xc, yc, zc,
0.5 * box_edge / unit.angstrom)
# Packmol setting for the ions section
if neutralize_solute and n_ions >= 1:
body += '\n\n# Counter Ions'
for i in range(0, len(ions_smiles)):
if geometry == 'box':
body += solvent_template.format(ions_smiles_pdbs[i], n_ions,
xmin, ymin, zmin, xmax, ymax,
zmax)
if geometry == 'sphere':
body += solvent_template.format(ions_smiles_pdbs[i], n_ions,
xc, yc, zc,
0.5 * box_edge / unit.angstrom)
# Packmol configuration file
packmol_filename = os.path.basename(tempfile.mktemp(suffix='.inp'))
with open(packmol_filename, 'w') as file_handle:
file_handle.write(body)
# Call Packmol
if not verbose:
mute_output = open(os.devnull, 'w')
with open(packmol_filename, 'r') as file_handle:
subprocess.check_call(['packmol'],
stdin=file_handle,
stdout=mute_output,
stderr=mute_output)
else:
with open(packmol_filename, 'r') as file_handle:
subprocess.check_call(['packmol'], stdin=file_handle)
# Read in the Packmol solvated system
solvated = oechem.OEMol()
if os.path.exists(mixture_pdb + '_FORCED'):
os.rename(mixture_pdb + '_FORCED', mixture_pdb)
print("Warning: Packing solution is not optimal")
ifs = oechem.oemolistream(mixture_pdb)
oechem.OEReadMolecule(ifs, solvated)
# To avoid to change the user oemol starting solute by reading in
# the generated mixture pdb file and loosing molecule info, the
# solvent molecules are extracted from the mixture system and
# added back to the starting solute
# Extract from the solution system the solvent molecules
# by checking the previous solute generated ID: id+resname+chainID
hv_solvated = oechem.OEHierView(
solvated,
oechem.OEAssumption_BondedResidue + oechem.OEAssumption_ResPerceived)
# This molecule will hold the solvent molecules generated directly from
# the omega conformers. This is useful to avoid problems related to read in
# the solvent molecules from pdb files and triggering unwanted perceiving actions
new_components = oechem.OEMol()
bv = oechem.OEBitVector(solvated.GetMaxAtomIdx())
for chain in hv_solvated.GetChains():
for frag in chain.GetFragments():
for hres in frag.GetResidues():
oe_res = hres.GetOEResidue()
if str(oe_res.GetResidueNumber()) + oe_res.GetName(
) + chain.GetChainID() not in solute_resid_list:
oechem.OEAddMols(new_components,
solvent_resid_dic_names[oe_res.GetName()])
atms = hres.GetAtoms()
for at in atms:
bv.SetBitOn(at.GetIdx())
pred = oechem.OEAtomIdxSelected(bv)
components = oechem.OEMol()
oechem.OESubsetMol(components, solvated, pred)
new_components.SetCoords(components.GetCoords())
# This is necessary otherwise just one big residue is created
oechem.OEPerceiveResidues(new_components)
# Add the solvent molecules to the solute copy
solvated_system = solute.CreateCopy()
oechem.OEAddMols(solvated_system, new_components)
# Set Title
solvated_system.SetTitle(solute.GetTitle())
# Set ions resname to Na+ and Cl-
for at in solvated_system.GetAtoms():
res = oechem.OEAtomGetResidue(at)
if res.GetName() == ' NA':
res.SetName("Na+")
oechem.OEAtomSetResidue(atmol, res)
elif res.GetName() == ' CL':
res.SetName("Cl-")
oechem.OEAtomSetResidue(atmol, res)
else:
pass
# Cleaning
to_delete = solvent_pdbs + [packmol_filename, solute_pdb, mixture_pdb]
if salt_concentration > 0.0 * unit.millimolar and n_salt >= 1:
to_delete += salt_pdbs
if neutralize_solute and n_ions >= 1:
to_delete += ions_smiles_pdbs
for fn in to_delete:
try:
os.remove(fn)
except:
pass
# Calculate the solution total density
total_wgt = oechem.OECalculateMolecularWeight(
solvated_system) * unit.gram / unit.mole
density_mix = (1 / unit.AVOGADRO_CONSTANT_NA) * total_wgt / Volume
print("Computed Solution Density = {}".format(
density_mix.in_units_of(unit.gram / unit.milliliter)))
# Threshold checking
ths = 0.1 * unit.gram / unit.milliliter
if not abs(density -
density_mix.in_units_of(unit.gram / unit.milliliter)) < ths:
raise ValueError(
"Error: the computed density for the solute {} does not match the selected density {} vs {}"
.format(solute.GetTitle(), density_mix, density))
if geometry == 'box':
# Define the box vector and attached it as SD tag to the solvated system
# with ID tag: 'box_vectors'
box_vectors = (Vec3(box_edge / unit.angstrom, 0.0,
0.0), Vec3(0.0, box_edge / unit.angstrom, 0.0),
Vec3(0.0, 0.0,
box_edge / unit.angstrom)) * unit.angstrom
box_vectors = data_utils.encodePyObj(box_vectors)
solvated_system.SetData(oechem.OEGetTag('box_vectors'), box_vectors)
if return_components:
new_components.SetTitle(solute.GetTitle() + '_solvent_comp')
return solvated_system, new_components
else:
return solvated_system
def PrepareReceptor(pdb,padding=4,outpath=""):
"""
Prepares a receptor from a pdb with a crystalized ligand
Padding controls the docking region.
If outpath is given, PrepareReceptor will write an openeye binary (oeb) of the receptor structure. This will be faster than rebuilding the receptor every time.
"""
print("STOP CALLING THIS FUNCTION")
exit()
com = oechem.OEGraphMol()
ifs = oechem.oemolistream()
if ifs.open(pdb):
oechem.OEReadPDBFile(ifs, com)
ifs.close()
"""
Sorry, this requires some explanation. Openeye wasn't recognizing the previously docked ligand, so I tried to find other ways.
The next blocks of code take our system and split it based on its connected components, for which its REQUIRED that our protein
only has a single chain. It assumes that the last component is the ligand. It then creates the ligand (lig) and protein (prot)
as separate molecules. Next, it finds the minimum and maximum 3D coordinates of the current ligand and produces a box around
it with the specified padding. Finally it uses this box to create a 'receptor' object into which ligands can be docked.
Only the receptor is returned.
Openeye's docking shouldn't be this involved, but I couldn't get it to run the typical 'hybrid' docking without error.
"""
oechem.OEDetermineConnectivity(com)
nparts, connect = oechem.OEDetermineComponents(com)
if(nparts != 2):
print("ERR in dock_conf::prepareReceptor. PDB doesn't have 2 connected components")
exit()
## TODO: What is a good way to catch errors?
# Get apo
pred = oechem.OEPartPredAtom(connect)
pred.SelectPart(nparts)
lig = oechem.OEGraphMol()
oechem.OESubsetMol(lig, com, pred)
print(lig)
# Get protein
pred = oechem.OEPartPredAtom(connect)
pred.SelectPart(1)
prot = oechem.OEGraphMol()
oechem.OESubsetMol(prot, com, pred)
# Get box dimensions by iterating over ligand
x_min = y_min = z_min = float('inf')
x_max = y_max = z_max = -float('inf')
crd = lig.GetCoords()
print("CRD", crd)
for atm in crd:
x,y,z = crd[atm]
if x < x_min:
x_min = x
if y < y_min:
y_min = y
if z < z_min:
z_min = z
if x > x_max:
x_max = x
if y > y_max:
y_max = y
if z > z_max:
z_max = z
x_min -= padding
y_min -= padding
z_min -= padding
x_max += padding
y_max += padding
z_max += padding
print(x_min,y_min,z_max, y_max)
# Now prepare the receptor
receptor = oechem.OEGraphMol()
box = oedocking.OEBox()
box.Setup(x_max, y_max, z_max, x_min, y_min, z_min)
oedocking.OEMakeReceptor(receptor, prot, box)
if not outpath == "":
oedocking.OEWriteReceptorFile(receptor,f'{outpath}/receptor.oeb')
return receptor
def _extract_oe_fragment(
molecule: Molecule, atom_indices: Set[int], bond_indices: Set[Tuple[int, int]]
) -> Molecule:
from openeye import oechem
oe_molecule = molecule.to_openeye()
# Restore the map indices as to_openeye does not automatically add them.
for atom_index, map_index in molecule.properties["atom_map"].items():
oe_atom = oe_molecule.GetAtom(oechem.OEHasAtomIdx(atom_index))
oe_atom.SetMapIdx(map_index)
# Include any Hs bonded to the included atom set so we can retain their map
# indices.
for map_index in {*atom_indices}:
oe_atom = oe_molecule.GetAtom(oechem.OEHasMapIdx(map_index))
for neighbour in oe_atom.GetAtoms():
if (
neighbour.GetAtomicNum() != 1
or neighbour.GetMapIdx() < 1
or neighbour.GetMapIdx() in atom_indices
):
continue
atom_indices.add(neighbour.GetMapIdx())
bond_indices.add((map_index, neighbour.GetMapIdx()))
atom_bond_set = oechem.OEAtomBondSet()
for map_index in atom_indices:
atom = oe_molecule.GetAtom(oechem.OEHasMapIdx(map_index))
atom_bond_set.AddAtom(atom)
for map_index_1, map_index_2 in bond_indices:
atom_1 = oe_molecule.GetAtom(oechem.OEHasMapIdx(map_index_1))
atom_2 = oe_molecule.GetAtom(oechem.OEHasMapIdx(map_index_2))
bond = oe_molecule.GetBond(atom_1, atom_2)
if not bond:
raise ValueError(f"{(map_index_1, map_index_2)} is a disconnected bond")
atom_bond_set.AddBond(bond)
atom_predicate = oechem.OEIsAtomMember(atom_bond_set.GetAtoms())
bond_predicate = oechem.OEIsBondMember(atom_bond_set.GetBonds())
fragment = oechem.OEMol()
oechem.OESubsetMol(fragment, oe_molecule, atom_predicate, bond_predicate, True)
oechem.OEAddExplicitHydrogens(fragment)
oechem.OEPerceiveChiral(fragment)
# Always restore map?
# if restore_maps:
# In some cases (symmetric molecules) this changes the atom map so skip it
# restore_atom_map(fragment)
# atom map should be restored for combinatorial fragmentation
# Perceive stereo and check that defined stereo did not change
oechem.OEPerceiveChiral(fragment)
oechem.OE3DToAtomStereo(fragment)
oechem.OE3DToBondStereo(fragment)
return Molecule.from_openeye(fragment, allow_undefined_stereo=True)
pattern = oechem.OEGraphMol()
target = oechem.OEGraphMol()
oechem.OESmilesToMol(pattern, "c1(cc(nc2c1C(CCC2)Cl)CCl)O")
oechem.OESmilesToMol(target, "c1(c2c(nc(n1)CF)COC=C2)N")
# @ <SNIPPET-EXPR>
atomexpr = oechem.OEExprOpts_DefaultAtoms
bondexpr = oechem.OEExprOpts_DefaultBonds
# @ </SNIPPET-EXPR>
patternQ = oechem.OEQMol(pattern)
# generate query with atom and bond expression options
# @ <SNIPPET-BUILDEXPR>
patternQ.BuildExpressions(atomexpr, bondexpr)
# @ </SNIPPET-BUILDEXPR>
mcss = oechem.OEMCSSearch(patternQ)
unique = True
count = 1
# loop over matches
for match in mcss.Match(target, unique):
print("Match %d:" % count)
print("Number of matched atoms: %d" % match.NumAtoms())
print("Number of matched bonds: %d" % match.NumBonds())
# create match subgraph
m = oechem.OEGraphMol()
oechem.OESubsetMol(m, match, True)
print("match smiles = %s" % oechem.OEMolToSmiles(m))
count += 1
# @ </SNIPPET>
# TERMS FOR USE OF SAMPLE CODE The software below ("Sample Code") is
# provided to current licensees or subscribers of OpenEye products or
# SaaS offerings (each a "Customer").
# Customer is hereby permitted to use, copy, and modify the Sample Code,
# subject to these terms. OpenEye claims no rights to Customer's
# modifications. Modification of Sample Code is at Customer's sole and
# exclusive risk. Sample Code may require Customer to have a then
# current license or subscription to the applicable OpenEye offering.
# THE SAMPLE CODE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
# EXPRESS OR IMPLIED. OPENEYE DISCLAIMS ALL WARRANTIES, INCLUDING, BUT
# NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
# PARTICULAR PURPOSE AND NONINFRINGEMENT. In no event shall OpenEye be
# liable for any damages or liability in connection with the Sample Code
# or its use.
# @ <SNIPPET>
from __future__ import print_function
from openeye import oechem
from openeye import oemedchem
mol = oechem.OEGraphMol()
oechem.OESmilesToMol(mol, "CCOc1ccc(cc1)CC(OC)c2ccccc2CC(=O)N")
adjustHCount = True
for frag in oemedchem.OEGetBemisMurcko(mol):
fragment = oechem.OEGraphMol()
oechem.OESubsetMol(fragment, mol, frag, adjustHCount)
print(".".join(r.GetName() for r in frag.GetRoles()),
oechem.OEMolToSmiles(fragment))
# @ </SNIPPET>
def main(argv=[__name__]):
itf = oechem.OEInterface(InterfaceData, argv)
# flag on command line indicates uncoloring option or not
bUncolor = itf.GetBool("-uncolor")
# input structure(s) to transform
ifsmols = oechem.oemolistream()
if not ifsmols.open(itf.GetString("-i")):
oechem.OEThrow.Fatal("Unable to open %s for reading" %
itf.GetString("-i"))
# save output structure(s) to this file
ofs = oechem.oemolostream()
if not ofs.open(itf.GetString("-o")):
oechem.OEThrow.Fatal("Unable to open %s for writing" %
itf.GetString("-o"))
if not oechem.OEIsSDDataFormat(ofs.GetFormat()):
oechem.OEThrow.Fatal("Unable to open %s for writing" %
itf.GetString("-o"))
irec = 0
ototal = 0
frag = oechem.OEGraphMol()
for mol in ifsmols.GetOEGraphMols():
irec += 1
oechem.OEDeleteEverythingExceptTheFirstLargestComponent(mol)
iter = oemedchem.OEGetBemisMurcko(mol)
if not iter.IsValid():
name = mol.GetTitle()
if not mol.GetTitle():
name = 'Record ' + str(irec)
oechem.OEThrow.Warning("%s: no perceived regions" % name)
continue
for bmregion in iter:
# create a fragment from the perceived region
oechem.OESubsetMol(frag, mol, bmregion, True)
if bUncolor:
# ignore 3D stereo parities
if (frag.GetDimension() == 3):
frag.SetDimension(0)
# uncolor the fragment
oechem.OEUncolorMol(frag)
smi = oechem.OEMolToSmiles(frag)
# annotate the input molecule with the role information
for role in bmregion.GetRoles():
oechem.OEAddSDData(mol, role.GetName(), smi)
ototal += 1
oechem.OEWriteMolecule(ofs, mol)
if not irec:
oechem.OEThrow.Fatal('No records in input structure file to perceive')
if not ototal:
oechem.OEThrow.Warning('No annotated structures generated')
print(
"Input molecules={0:d}, output annotated {1:s}molecules={2:d}".format(
irec, ("(uncolored) " if bUncolor else ""), ototal))
return 0