def _build_fragment_from_complex(complex,
residue,
ligand,
ligand_core,
result=0,
substructure=True,
simmetry=False):
from rdkit import Chem
import rdkit.Chem.rdmolops as rd
import rdkit.Chem.rdchem as rc
import rdkit.Chem.AllChem as rp
#Retrieve atom core linking fragment
try:
atom_core_idx, atoms_core, _ = _search_core_fragment_linker(
ligand, ligand_core, result, simmetry)
except TypeError:
raise ce.SameMolecule(
"Core and ligand are the exact same molecule. Check your inputs")
atom_core = at.Atom(ligand_core, atom_core_idx)
mol = Chem.MolFromPDBFile(complex, removeHs=False)
#Retrieve hydrogen core attach to linking atom
original = rd.SplitMolByPDBResidues(mol)[residue]
hydrogen_core_idx = [
atom.GetIdx()
for atom in original.GetAtomWithIdx(atom_core_idx).GetNeighbors()
if atom.GetAtomicNum() == 1
][0]
hydrogen_core = at.Atom(original, hydrogen_core_idx)
# Delete core for full ligand with substructure
# and if it fails manually
#Chem.MolToPDBFile(ligand, "int0.pdb")
if substructure:
fragment = rd.DeleteSubstructs(ligand, ligand_core)
new_mol = rc.EditableMol(fragment)
for atom in reversed(fragment.GetAtoms()):
neighbours = atom.GetNeighbors()
if len(neighbours) == 0: new_mol.RemoveAtom(atom.GetIdx())
else:
new_mol = rc.EditableMol(ligand)
for atom in atoms_core:
new_mol.RemoveAtom(atom)
#Chem.MolToPDBFile(new_mol.GetMol(), "int1.pdb")
for atom in reversed(new_mol.GetMol().GetAtoms()):
neighbours = atom.GetNeighbors()
if len(neighbours) == 0: new_mol.RemoveAtom(atom.GetIdx())
#Add missing hydrogen to full ligand and create pdb differently
#depending on the previous step
fragment = new_mol.GetMol()
#Chem.MolToPDBFile(fragment, "int2.pdb")
old_atoms = [atom.GetIdx() for atom in fragment.GetAtoms()]
if substructure:
fragment = Chem.AddHs(fragment, False, True)
else:
#res = rp.EmbedMolecule(fragment)
fragment = Chem.AddHs(fragment, False, True)
return fragment, old_atoms, hydrogen_core, atom_core
def _build_fragment_from_complex(complex, residue, ligand, ligand_core, result=0, substructure=True, simmetry=False):
from rdkit import Chem
import rdkit.Chem.rdmolops as rd
import rdkit.Chem.rdchem as rc
# Retrieve atom core linking fragment
try:
atom_core_idx, atoms_core, atom_fragment = _search_core_fragment_linker(ligand, ligand_core, result, simmetry)
print("ATOM OF FRAGMENT ATTACHED TO CORE:", atom_fragment)
print("ATOM OF CORE ATTACHED TO FRAGMENT:", atom_core_idx)
except TypeError:
raise ce.SameMolecule("Core and ligand are the exact same molecule. Check your inputs.")
atom_core = at.Atom(ligand_core, atom_core_idx)
mol = Chem.MolFromPDBFile(complex, removeHs=False)
# Retrieve hydrogen core attach to linking atom
original = rd.SplitMolByPDBResidues(mol)[residue]
hydrogen_core_idx = \
[atom.GetIdx() for atom in original.GetAtomWithIdx(atom_core_idx).GetNeighbors() if atom.GetAtomicNum() == 1][0]
hydrogen_core = at.Atom(original, hydrogen_core_idx)
# Delete core for full ligand with substructure and if it fails manually
if substructure:
fragment = rd.DeleteSubstructs(ligand, ligand_core)
new_mol = rc.EditableMol(fragment)
for atom in reversed(fragment.GetAtoms()):
neighbours = atom.GetNeighbors()
if len(neighbours) == 0 and atom.GetAtomicNum() == 1:
new_mol.RemoveAtom(atom.GetIdx())
else:
new_mol = rc.EditableMol(ligand)
for atom in reversed(atoms_core):
new_mol.RemoveAtom(atom)
for atom in reversed(new_mol.GetMol().GetAtoms()):
neighbours = atom.GetNeighbors()
if len(neighbours) == 0 and atom.GetAtomicNum() == 1:
new_mol.RemoveAtom(atom.GetIdx())
print("FRAGMENT ATOMS", [atom.GetIdx() for atom in new_mol.GetMol().GetAtoms()])
# Add missing hydrogen to full ligand and create pdb differently depending on the previous step
fragment = new_mol.GetMol()
old_atoms = [atom.GetIdx() for atom in fragment.GetAtoms() if atom.GetAtomicNum() != 1]
new_atoms = [atom.GetIdx() for atom in ligand.GetAtoms() if
atom.GetIdx() not in atoms_core and atom.GetAtomicNum() != 1]
print("old_atoms", old_atoms, "new_atoms", new_atoms)
mapping = {new_atom: old_atom for new_atom, old_atom in zip(new_atoms, old_atoms)}
atom_fragment_mapped = mapping[atom_fragment]
fragment = Chem.AddHs(fragment, False, True)
correct = _check_fragment(fragment, ligand, mapping, atom_fragment, atom_fragment_mapped, ligand_core)
Chem.MolToPDBFile(fragment, "int0.pdb")
assert len(old_atoms) == len(new_atoms)
return fragment, old_atoms, hydrogen_core, atom_core, atom_fragment, mapping, correct
def to_rdkit_Mol(mol):
rdmol = rdchem.EditableMol(rdchem.Mol())
for atom in mol.atoms:
rdmol.AddAtom(rdchem.Atom(Element(atom).Z))
for bond in mol.bondmap:
rdmol.AddBond(bond[0], bond[1], rdchem.BondType.values[bond[2]])
return rdmol.GetMol()
def attach_capping(mol1, mol2):
"""it is connecting all Nterminals with the desired capping
Arguments:
mol1 {rdKit mol object} -- first molecule to be connected
mol2 {rdKit mol object} -- second molecule to be connected - chosen N-capping
Returns:
rdKit mol object -- mol1 updated (connected with mol2, one or more)
"""
count = 0
# detects all the N terminals in mol1
for atom in mol1.GetAtoms():
atom.SetProp('Cterm', 'False')
if atom.GetSmarts() == '[N:2]' or atom.GetSmarts(
) == '[NH2:2]' or atom.GetSmarts() == '[NH:2]':
count += 1
atom.SetProp('Nterm', 'True')
else:
atom.SetProp('Nterm', 'False')
# detects all the C terminals in mol2 (it should be one)
for atom in mol2.GetAtoms():
atom.SetProp('Nterm', 'False')
if atom.GetSmarts() == '[C:1]' or atom.GetSmarts() == '[CH:1]':
atom.SetProp('Cterm', 'True')
else:
atom.SetProp('Cterm', 'False')
# mol2 is addes to all the N terminal of mol1
for i in range(count):
combo = rdmolops.CombineMols(mol1, mol2)
Nterm = []
Cterm = []
# saves in two different lists the index of the atoms which has to be connected
for atom in combo.GetAtoms():
if atom.GetProp('Nterm') == 'True':
Nterm.append(atom.GetIdx())
if atom.GetProp('Cterm') == 'True':
Cterm.append(atom.GetIdx())
# creates the amide bond
edcombo = rdchem.EditableMol(combo)
edcombo.AddBond(Nterm[0], Cterm[0], order=Chem.rdchem.BondType.SINGLE)
clippedMol = edcombo.GetMol()
# removes tags and lables form the atoms which reacted
clippedMol.GetAtomWithIdx(Nterm[0]).SetProp('Nterm', 'False')
clippedMol.GetAtomWithIdx(Cterm[0]).SetProp('Cterm', 'False')
clippedMol.GetAtomWithIdx(Nterm[0]).SetAtomMapNum(0)
clippedMol.GetAtomWithIdx(Cterm[0]).SetAtomMapNum(0)
# uptades the 'core' molecule
mol1 = clippedMol
return mol1
def cyclize(mol, cy):
"""it is connecting cyclizing the given molecule
Arguments:
mol {rdKit mol object} -- molecule to be cyclized
cy {int} -- 1=yes, 0=no cyclazation
Returns:
mols {list of rdKit mol objects} -- possible cyclazation
"""
count = 0
# detects all the N terminals in mol
for atom in mol.GetAtoms():
if atom.GetSmarts() == '[N:2]' or atom.GetSmarts(
) == '[NH2:2]' or atom.GetSmarts() == '[NH:2]':
count += 1
atom.SetProp('Nterm', 'True')
else:
atom.SetProp('Nterm', 'False')
# detects all the C terminals in mol (it should be one)
for atom in mol.GetAtoms():
if atom.GetSmarts() == '[C:1]' or atom.GetSmarts() == '[CH:1]':
atom.SetProp('Cterm', 'True')
else:
atom.SetProp('Cterm', 'False')
# detects all the S terminals in mol
for atom in mol.GetAtoms():
if atom.GetSmarts() == '[S:1]':
atom.SetProp('Sact1', 'True')
else:
atom.SetProp('Sact1', 'False')
for atom in mol.GetAtoms():
if atom.GetSmarts() == '[S:2]':
atom.SetProp('Sact2', 'True')
else:
atom.SetProp('Sact2', 'False')
for atom in mol.GetAtoms():
if atom.GetSmarts() == '[S:3]':
atom.SetProp('Sact3', 'True')
else:
atom.SetProp('Sact3', 'False')
Nterm = []
Cterm = []
Sact1 = []
Sact2 = []
Sact3 = []
# saves active Cysteins postions:
for atom in mol.GetAtoms():
if atom.GetProp('Sact1') == 'True':
Sact1.append(atom.GetIdx())
# saves active Cysteins 2 postions:
for atom in mol.GetAtoms():
if atom.GetProp('Sact2') == 'True':
Sact2.append(atom.GetIdx())
# saves active Cysteins 3 postions:
for atom in mol.GetAtoms():
if atom.GetProp('Sact3') == 'True':
Sact3.append(atom.GetIdx())
# creates the S-S bond (in the current version only two 'active' Cys, this codo picks two random anyway):
while len(Sact1) >= 2:
edmol = rdchem.EditableMol(mol)
pos = list(range(len(Sact1)))
x = np.random.choice(pos, 1)[0]
pos.remove(x)
y = np.random.choice(pos, 1)[0]
a = Sact1[x]
b = Sact1[y]
edmol.AddBond(a, b, order=Chem.rdchem.BondType.SINGLE)
mol = edmol.GetMol()
mol.GetAtomWithIdx(a).SetProp('Sact1', 'False')
mol.GetAtomWithIdx(b).SetProp('Sact1', 'False')
mol.GetAtomWithIdx(a).SetAtomMapNum(0)
mol.GetAtomWithIdx(b).SetAtomMapNum(0)
Sact1.remove(a)
Sact1.remove(b)
while len(Sact2) >= 2:
edmol = rdchem.EditableMol(mol)
pos = list(range(len(Sact2)))
x = np.random.choice(pos, 1)[0]
pos.remove(x)
y = np.random.choice(pos, 1)[0]
a = Sact2[x]
b = Sact2[y]
edmol.AddBond(a, b, order=Chem.rdchem.BondType.SINGLE)
mol = edmol.GetMol()
mol.GetAtomWithIdx(a).SetProp('Sact2', 'False')
mol.GetAtomWithIdx(b).SetProp('Sact2', 'False')
mol.GetAtomWithIdx(a).SetAtomMapNum(0)
mol.GetAtomWithIdx(b).SetAtomMapNum(0)
Sact2.remove(a)
Sact2.remove(b)
while len(Sact3) >= 2:
edmol = rdchem.EditableMol(mol)
pos = list(range(len(Sact3)))
x = np.random.choice(pos, 1)[0]
pos.remove(x)
y = np.random.choice(pos, 1)[0]
a = Sact3[x]
b = Sact3[y]
edmol.AddBond(a, b, order=Chem.rdchem.BondType.SINGLE)
mol = edmol.GetMol()
mol.GetAtomWithIdx(a).SetProp('Sact3', 'False')
mol.GetAtomWithIdx(b).SetProp('Sact3', 'False')
mol.GetAtomWithIdx(a).SetAtomMapNum(0)
mol.GetAtomWithIdx(b).SetAtomMapNum(0)
Sact3.remove(a)
Sact3.remove(b)
# saves active C and N terminals postions:
for atom in mol.GetAtoms():
if atom.GetProp('Nterm') == 'True':
Nterm.append(atom.GetIdx())
if atom.GetProp('Cterm') == 'True':
Cterm.append(atom.GetIdx())
if cy == 1:
edmol = rdchem.EditableMol(mol)
# creates the amide bond
edmol.AddBond(Nterm[0], Cterm[0], order=Chem.rdchem.BondType.SINGLE)
edmol.RemoveAtom(Cterm[0] + 1)
mol = edmol.GetMol()
# removes tags and lables form the atoms which reacted
mol.GetAtomWithIdx(Nterm[0]).SetProp('Nterm', 'False')
mol.GetAtomWithIdx(Cterm[0]).SetProp('Cterm', 'False')
mol.GetAtomWithIdx(Nterm[0]).SetAtomMapNum(0)
mol.GetAtomWithIdx(Cterm[0]).SetAtomMapNum(0)
return mol
def connect_mol(mol1, mol2):
"""it is connecting all Nterminals of mol1 with the Cterminal
of the maximum possible number of mol2s
Arguments:
mol1 {rdKit mol object} -- first molecule to be connected
mol2 {rdKit mol object} -- second molecule to be connected
Returns:
rdKit mol object -- mol1 updated (connected with mol2, one or more)
"""
# used internally to recognize a methylated aa:
metbond = False
# can be set with exclude or allow methylation,
# it refers to the possibility of having methylation in the entire GA:
methyl = False
count = 0
# detects all the N terminals in mol1
for atom in mol1.GetAtoms():
atom.SetProp('Cterm', 'False')
atom.SetProp('methyl', 'False')
if atom.GetSmarts() == '[N:2]' or atom.GetSmarts(
) == '[NH2:2]' or atom.GetSmarts() == '[NH:2]':
count += 1
atom.SetProp('Nterm', 'True')
else:
atom.SetProp('Nterm', 'False')
# detects all the C terminals in mol2 (it should be one)
for atom in mol2.GetAtoms():
atom.SetProp('Nterm', 'False')
atom.SetProp('methyl', 'False')
if atom.GetSmarts() == '[C:1]' or atom.GetSmarts() == '[CH:1]':
atom.SetProp('Cterm', 'True')
else:
atom.SetProp('Cterm', 'False')
# mol2 is addes to all the N terminal of mol1
for i in range(count):
combo = rdmolops.CombineMols(mol1, mol2)
Nterm = []
Cterm = []
# saves in two different lists the index of the atoms which has to be connected
for atom in combo.GetAtoms():
if atom.GetProp('Nterm') == 'True':
Nterm.append(atom.GetIdx())
if atom.GetProp('Cterm') == 'True':
Cterm.append(atom.GetIdx())
# creates the amide bond
edcombo = rdchem.EditableMol(combo)
edcombo.AddBond(Nterm[0], Cterm[0], order=Chem.rdchem.BondType.SINGLE)
edcombo.RemoveAtom(Cterm[0] + 1)
clippedMol = edcombo.GetMol()
# removes tags and lables form c term atoms which reacted
clippedMol.GetAtomWithIdx(Cterm[0]).SetProp('Cterm', 'False')
clippedMol.GetAtomWithIdx(Cterm[0]).SetAtomMapNum(0)
# methylates amide bond
if metbond == True and methyl == True:
Nterm = []
Met = []
methyl = rdmolfiles.MolFromSmiles('[C:4]')
for atom in methyl.GetAtoms():
atom.SetProp('methyl', 'True')
atom.SetProp('Nterm', 'False')
atom.SetProp('Cterm', 'False')
metcombo = rdmolops.CombineMols(clippedMol, methyl)
for atom in metcombo.GetAtoms():
if atom.GetProp('Nterm') == 'True':
Nterm.append(atom.GetIdx())
if atom.GetProp('methyl') == 'True':
Met.append(atom.GetIdx())
metedcombo = rdchem.EditableMol(metcombo)
metedcombo.AddBond(Nterm[0],
Met[0],
order=Chem.rdchem.BondType.SINGLE)
clippedMol = metedcombo.GetMol()
clippedMol.GetAtomWithIdx(Met[0]).SetProp('methyl', 'False')
clippedMol.GetAtomWithIdx(Met[0]).SetAtomMapNum(0)
# removes tags and lables form the atoms which reacted
clippedMol.GetAtomWithIdx(Nterm[0]).SetProp('Nterm', 'False')
clippedMol.GetAtomWithIdx(Nterm[0]).SetAtomMapNum(0)
# uptades the 'core' molecule
mol1 = clippedMol
metbond = False
return mol1
def _build_fragment_from_complex(complex,
residue,
ligand,
ligand_core,
result=0,
substructure=True,
symmetry=False,
frag_core_atom=None):
"""
Parameters
----------
complex : str
Path of PDB file with protein-ligand complex.
residue : str
Residue name.
ligand : RDKit molecule object
Ligand to grow during the simulation.
ligand_core : RDKit molecule object
Common structure of each grown ligand.
result : int
Index to extract core atoms from the substructure search.
substructure : bool
Delete core for full ligand with substructure.
symmetry : bool
Check the symmetry of the ligand.
Returns
-------
fragment : RDKit molecule object
Grown ligand with deleted core atoms.
old_atoms : list
Idx for each atom of grown ligand without core.
hydrogen_core : pele_platform.Frag.atoms.Atom
Hydrogen core attached to linking atom.
atom_core : pele_platform.Frag.atoms.Atom
Atom object of the hydrogen core atom attached to linking atom.
atom_fragment : int
Atom of the fragment attached to the core.
mapping : dict
Mapping of Idx for old atoms and full fragment atoms.
correct : bool
Checks if the fragment is correct (fragment size, chirality, missing hydrogens)
Raises
------
TypeError
If the core and ligand are the same molecule.
"""
# Retrieve atom core linking fragment
try:
atom_core_idx, atoms_core, atom_fragment = _search_core_fragment_linker(
ligand, ligand_core, result, symmetry, frag_core_atom)
print("ATOM OF FRAGMENT ATTACHED TO CORE:", atom_fragment)
print("ATOM OF CORE ATTACHED TO FRAGMENT:", atom_core_idx)
except TypeError:
raise ce.SameMolecule(
"Core and ligand are the exact same molecule. Check your inputs.")
atom_core = at.Atom(ligand_core, atom_core_idx)
mol = Chem.MolFromPDBFile(complex, removeHs=False)
# Retrieve hydrogen core attach to linking atom
original = rd.SplitMolByPDBResidues(mol)[residue]
hydrogen_core_idx = \
[atom.GetIdx() for atom in original.GetAtomWithIdx(atom_core_idx).GetNeighbors() if atom.GetAtomicNum() == 1][0]
hydrogen_core = at.Atom(original, hydrogen_core_idx)
# Delete core for full ligand with substructure and if it fails manually
if substructure:
fragment = rd.DeleteSubstructs(ligand, ligand_core)
new_mol = rc.EditableMol(fragment)
for atom in reversed(fragment.GetAtoms()):
neighbours = atom.GetNeighbors()
if len(neighbours) == 0 and atom.GetAtomicNum() == 1:
new_mol.RemoveAtom(atom.GetIdx())
else:
new_mol = rc.EditableMol(ligand)
for atom in reversed(atoms_core):
new_mol.RemoveAtom(atom)
for atom in reversed(new_mol.GetMol().GetAtoms()):
neighbours = atom.GetNeighbors()
if len(neighbours) == 0 and atom.GetAtomicNum() == 1:
new_mol.RemoveAtom(atom.GetIdx())
print("FRAGMENT ATOMS",
[atom.GetIdx() for atom in new_mol.GetMol().GetAtoms()])
# Add missing hydrogen to full ligand and create pdb differently depending on the previous step
fragment = new_mol.GetMol()
old_atoms = [
atom.GetIdx() for atom in fragment.GetAtoms()
if atom.GetAtomicNum() != 1
]
new_atoms = [
atom.GetIdx() for atom in ligand.GetAtoms()
if atom.GetIdx() not in atoms_core and atom.GetAtomicNum() != 1
]
print("old_atoms", old_atoms, "new_atoms", new_atoms)
mapping = {
new_atom: old_atom
for new_atom, old_atom in zip(new_atoms, old_atoms)
}
atom_fragment_mapped = mapping[atom_fragment]
# Retrieve all atom idxs before adding hydrogen to fragment_atom
fragment_atoms_wo_hydrogen = [
atom.GetIdx() for atom in fragment.GetAtoms()
]
fragment = Chem.AddHs(fragment, False, True)
correct = _check_fragment(fragment, ligand, mapping, atom_fragment,
atom_fragment_mapped, ligand_core)
Chem.MolToPDBFile(fragment, "int0.pdb")
assert len(old_atoms) == len(new_atoms)
return fragment, old_atoms, hydrogen_core, atom_core, atom_fragment, mapping, correct, fragment_atoms_wo_hydrogen
def FragmentMol(query_mol, query_mcs_indices):
answer = rdchem.EditableMol(query_mol)
all_query_bonds = query_mol.GetBonds()
for bond_index in range(len(query_mol.GetBonds()):
begin_atom_idx = all_query_bonds[bond_index].GetBeginAtom().GetIdx()
end_atom_idx = all_query_bonds[bond_index].GetEndAtom().GetIdx()
# if the bond is between mcs core and a side-chain - remove it
rdchem.EditableMol.RemoveBond(answer, begin_atom_idx, end_atom_idx) if ( bool(begin_atom_idx in query_mcs_indices) != bool(end_atom_idx in query_mcs_indices) ) else None
return(answer.GetMol())
def MapIndexQ2T(query_index, query_mcs_indices, template_mcs_indices):
# get index of query_index in query_mcs_indices
# return template_mcs_indices[index]
# else return -1
class Sidechain:
def __init__():
anchor = (-1, -1, -1) # path of length 3 starting in atom where sidechain connects that goes through 2 other atoms that are part of a skeleton
# these are stored not as indices in context of query mol but instead as order in query_mcs_indices
# it is assumed this order is identical in template_mcs_indices and thus is used for accessing atom indices directly in template_mol
members = list() # contains indices of atoms (in context of whole molecule) that belong to this sidechain
build_order = list() # indices of members (in context of sidechain) listed in ascending order of distance from skeleton contact; ties are undefined but irrelevan
zpath = list() # list of triplets - first three indices (in context of whole molecule) of path to skeleton contact
zvals = list() # list of triplets - distance/angle/dihedral to atoms from zpath
dists_to_anchor = list() # distance of shortest path to skeleton contact - used for filling build_order
atom_type = list()
# add atom to sidechain by adding its index to members, first three steps of a path to anchor for z-matrix like reconstruction,
# zvals for distance+angle+dihedral and total distance to anchor
def AddAtom(query_mol, index):
# fill anchor values by providing index of a skeleton contact
def Anchor(query_mol, skeleton_contact, query_mcs_indices):
for mcs_idx_it in query_mcs_indices:
path = Chem.rdmolops.GetShortestPath(query_mol, skeleton_contact, mcs_idx_it)
if ( len(path) == 3 & path[1] in query_mcs_indices & path[2] in query_mcs_indices ):
anchor[0] = np.where(np.array(path[0]) == query_mcs_indices) # skeleton_contact
anchor[1] = np.where(np.array(path[1]) == query_mcs_indices) # way_to_anchor
anchor[2] = np.where(np.array(path[2]) == query_mcs_indices) # anchor
return(0)
return(1)
# fill build_order
def OrderUp(query_mol):
for it in members:
path = Chem.rdmolops.GetShortestPath(query_mol, it, self.skeleton_contact) # path will include starting and ending index
dists_to_anchor.add(len(path)) # remember the length
# we pay special attention if the path is not long enough to build the atom from zvals of this sidechain only
if (len(path)) == 2 ):
# atom is adjacent to skeleton contact
# dihedral and angle from template is used; only distance is taken from query
if (len(path)) == 3 ):
# atom is 2 steps away from skeleton contact
# only dihedral from template is used - IS THIS WELL DEFINED? LAST ATOM OF DIHEDRAL IS FROM QUERY, IS DIFF FROM SKELETON CONTACT AND IS AMBIGUOUS. DOES THIS MATTER?
# !! this should be ok if angles are not changed
# angle and distance are taken from query
if (len(path)) > 3 ):
# all zvals take from sidechain
zpath.add(path[-3:]) # add last three steps of the path
# measure distance it - path[1] and use it as zval[0]
# measure angle it - path[1] - path[2] and use it as zval[1]
# measure dihedral it - path[1] - path[2] - path[3] and use it as zval[2]
# now time to order
build_order = np.argsort(dists_to_anchor)