Spaces:
Sleeping
Sleeping
File size: 18,718 Bytes
9dd777e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 |
""" Adjusts amide and ester bonds in PROTAC substructures. """
from typing import Tuple, Dict
from rdkit import Chem
from protac_splitter.chemoinformatics import (
dummy2query,
canonize,
)
from protac_splitter.display_utils import display_mol
from protac_splitter.evaluation import check_reassembly
def adjust_amide_bond(
substruct: Chem.Mol,
linker: Chem.Mol,
substruct_attachment_id: int,
verbose: int = 0,
) -> Tuple[Chem.Mol, Chem.Mol]:
"""
Adjust the amide bond between the substruct and linker substructure.
Handles the case when neighboring atoms of the amide bond are dummy atoms, which represent attachment points.
The linker will be modified with the required additional atoms.
Args:
substruct: The substructure of the substruct (protein of interest) that contains the amide bond.
linker: The linker molecule that connects substruct to the E3 ligase.
substruct_attachment_id: The attachment point ID in the substruct substructure. E.g., 1 for the POI, as in "[*:1]".
Returns:
Tuple[Chem.Mol, Chem.Mol]: The adjusted substruct and linker molecules, in that order.
"""
# Pseudo-code of the algorithm:
"""
```python
# Check if the amide bond (N-C=O) is in the substructure
if "N-C(=O)" in substruct:
if neighbor("N-C(=O)") == "[*:substruct]":
# If the neighboring atom of the amide bond is a dummy atom, i.e., attachment point
mark_protac_as_wrong("[PROTAC]")
# Identify the bond to split, i.e., the nitrogen-carbon bond, and split
"[*:substruct]-[<optional neighboring atom>]-N-[*:tmp]", "[*:tmp]-C(=O)-[rest of the PROTAC]" = split_PROTAC_at("N-C")
"[Linker]-N-[*:tmp]" = join("[Linker]-[*:substruct]", "[*:substruct]-N-[*:tmp]")
rename_attachment_point("[*:tmp]-C(=O)-[rest of the PROTAC]")
rename_attachment_point("[Linker]-N-[*:tmp]")
elif neighbor(neighbor("N-C(=O)")) == "[*:substruct]":
# If the second neighbor of athe amide bond is a dummy atom, i.e., attachment point
mark_protac_as_wrong("[PROTAC]")
# Do as above
# Identify the bond to split, i.e., the nitrogen-carbon bond, and split
"[*:substruct]-N-[*:tmp]", "[*:tmp]-C(=O)-[rest of the PROTAC]" = split_PROTAC_at("N-C")
"[Linker]-N-[*:tmp]" = join("[Linker]-[*:substruct]", "[*:substruct]-N-[*:tmp]")
rename_attachment_point("[*:tmp]-C(=O)-[rest of the PROTAC]")
rename_attachment_point("[Linker]-N-[*:tmp]")
```
"""
# Convert dummy atoms in substruct to query atoms for substructure search
query_substruct = dummy2query(substruct)
# Identify amide bond (N-C=O) in substruct substructure
amide_pattern = Chem.MolFromSmarts("[NX3][CX3](=[OX1])")
amide_matches = query_substruct.GetSubstructMatches(amide_pattern, useChirality=True)
if not amide_matches:
return substruct, linker # No amide bond found, return the original substruct
side_atom = None
nitrogen_idx_found, carbonyl_idx_found = None, None
for match in amide_matches:
nitrogen_idx, carbonyl_idx = match[0], match[1]
nitrogen_atom = query_substruct.GetAtomWithIdx(nitrogen_idx)
carbonyl_atom = query_substruct.GetAtomWithIdx(carbonyl_idx)
for amide_atom in [nitrogen_atom, carbonyl_atom]:
# Check neighboring atoms for attachment points
# NOTE: The dummy atom representing an attachment point have atomic number 0
for neighbor in amide_atom.GetNeighbors():
if neighbor.GetAtomicNum() == 0:
nitrogen_idx_found = nitrogen_idx
carbonyl_idx_found = carbonyl_idx
side_atom = "N" if amide_atom == nitrogen_atom else "C"
break
# If previous search failed, check the neighbors of the neighboring
# atoms (second-order neighbors)
if nitrogen_idx_found is None or carbonyl_idx_found is None:
for neighbor in amide_atom.GetNeighbors():
for second_neighbor in neighbor.GetNeighbors():
if second_neighbor.GetIdx() == carbonyl_idx or second_neighbor.GetIdx() == nitrogen_idx:
continue # Skip the opposite atom from the amide bond
if second_neighbor.GetAtomicNum() == 0:
nitrogen_idx_found = nitrogen_idx
carbonyl_idx_found = carbonyl_idx
side_atom = "N" if amide_atom == nitrogen_atom else "C"
break
else:
break
if nitrogen_idx_found is None or carbonyl_idx_found is None or side_atom is None:
return substruct, linker
# Split the amide bond and adjust
dummy_label = 3
dummy_labels = [(dummy_label, dummy_label)] # The E3 and substruct will have 1 and 2, so we need a third one
amid_bond_idx = query_substruct.GetBondBetweenAtoms(nitrogen_idx_found, carbonyl_idx_found).GetIdx()
fragments = Chem.FragmentOnBonds(query_substruct, [amid_bond_idx], addDummies=True, dummyLabels=dummy_labels)
# Get the fragments resulting from bond breaking
try:
mol_frags = Chem.GetMolFrags(fragments, asMols=True, sanitizeFrags=True)
except Exception as e:
print(e)
return substruct, linker
# Identify the "[*:substruct][<optional neighboring atom>]N[3*]" fragment, the other one will be the "truncated" substruct
amide_fragment_pattern = Chem.MolFromSmarts(f"[*:{substruct_attachment_id}][{side_atom}][{dummy_label}*]")
amide_fragment = None
substruct_fixed = None
if verbose:
print(f'Attachment point: *:{substruct_attachment_id}')
print('Substruct:')
display_mol(substruct)
print('Linker:')
display_mol(linker)
for frag in mol_frags:
if frag.HasSubstructMatch(dummy2query(amide_fragment_pattern)):
amide_fragment = frag
if verbose:
print('Amide fragment:')
display_mol(frag)
else:
if verbose:
print('Substruct fragment:')
display_mol(frag)
substruct_fixed = frag
if amide_fragment is None or substruct_fixed is None:
return substruct, linker
# In order for the function to be used "on linkers", we need to make sure
# that the amide fragment contains the attachment point of the substruct.
# If not, there's nothing to do.
if f'[*:{substruct_attachment_id}]' not in Chem.MolToSmiles(amide_fragment, canonical=True):
return substruct, linker
# Rename the "[3*]" attachment point on the amide fragment to "[*:3]"
amide_fragment_smiles = Chem.MolToSmiles(amide_fragment, canonical=True)
amide_fragment_smiles = amide_fragment_smiles.replace(f'[{dummy_label}*]', f'[*:{dummy_label}]')
amide_fragment_smiles = canonize(amide_fragment_smiles)
amide_fragment = Chem.MolFromSmiles(amide_fragment_smiles)
# Use molzip to join the linker and the fragment at the original attachment point
linker_fixed = Chem.molzip(linker, amide_fragment)
# Rename the "[*:3]" attachment point back to the original attachment point on the linker
linker_fixed_smiles = Chem.MolToSmiles(linker_fixed, canonical=True)
linker_fixed_smiles = linker_fixed_smiles.replace(f'[*:{dummy_label}]', f'[*:{substruct_attachment_id}]')
linker_fixed_smiles = canonize(linker_fixed_smiles)
linker_fixed = Chem.MolFromSmiles(linker_fixed_smiles)
# Rename the "[3*]" attachment point back to the original attachment point on the substruct
substruct_fixed_smiles = Chem.MolToSmiles(substruct_fixed, canonical=True)
substruct_fixed_smiles = substruct_fixed_smiles.replace(f'[{dummy_label}*]', f'[*:{substruct_attachment_id}]')
substruct_fixed_smiles = canonize(substruct_fixed_smiles)
substruct_fixed = Chem.MolFromSmiles(substruct_fixed_smiles)
return substruct_fixed, linker_fixed
def adjust_amide_bonds_in_substructs(
substructs: Dict[str, str],
protac_smiles: str,
poi_attachment_id: int = 1,
e3_attachment_id: int = 2,
) -> Dict[str, str]:
""" Adjusts the amide bonds in the substructures of a PROTAC. Just a wrapper function to apply it to multiple substructures.
Args:
substructs: The substructures of the PROTAC. A dictionary of SMILES with keys 'poi', 'linker', and 'e3'.
protac_smiles: The SMILES of the PROTAC for checking reassembly.
Returns:
The updated substructures dictionary.
"""
poi_mol = Chem.MolFromSmiles(substructs['poi'])
e3_mol = Chem.MolFromSmiles(substructs['e3'])
linker_mol = Chem.MolFromSmiles(substructs['linker'])
# Fix the amide group on the POI ligand
poi_mol, linker_mol = adjust_amide_bond(poi_mol, linker_mol, poi_attachment_id)
poi_smiles = Chem.MolToSmiles(poi_mol, canonical=True)
linker_smiles = Chem.MolToSmiles(linker_mol, canonical=True)
e3_smiles = substructs['e3']
if not check_reassembly(protac_smiles, '.'.join([poi_smiles, linker_smiles, e3_smiles])):
return substructs
# Fix the amide group on the E3 binder
e3_mol, linker_mol = adjust_amide_bond(e3_mol, linker_mol, e3_attachment_id)
e3_smiles = Chem.MolToSmiles(e3_mol, canonical=True)
linker_smiles = Chem.MolToSmiles(linker_mol, canonical=True)
if not check_reassembly(protac_smiles, '.'.join([poi_smiles, linker_smiles, e3_smiles])):
return substructs
# Fix the amide group on the linker, E3 side
linker_mol, e3_mol = adjust_amide_bond(linker_mol, e3_mol, e3_attachment_id)
e3_smiles = Chem.MolToSmiles(e3_mol, canonical=True)
linker_smiles = Chem.MolToSmiles(linker_mol, canonical=True)
if not check_reassembly(protac_smiles, '.'.join([poi_smiles, linker_smiles, e3_smiles])):
return substructs
# Fix the amide group on the linker, POI side
linker_mol, poi_mol = adjust_amide_bond(linker_mol, poi_mol, poi_attachment_id)
poi_smiles = Chem.MolToSmiles(poi_mol, canonical=True)
linker_smiles = Chem.MolToSmiles(linker_mol, canonical=True)
if not check_reassembly(protac_smiles, '.'.join([poi_smiles, linker_smiles, e3_smiles])):
return substructs
substructs['poi'] = poi_smiles
substructs['e3'] = e3_smiles
substructs['linker'] = linker_smiles
return substructs
def adjust_ester_bond(
substruct: Chem.Mol,
linker: Chem.Mol,
substruct_attachment_id: int,
verbose: int = 0,
) -> Tuple[Chem.Mol, Chem.Mol]:
"""
Adjust the amide bond between the substruct and linker substructure.
Handles the case when neighboring atoms of the amide bond are dummy atoms, which represent attachment points.
Args:
substruct: The substructure of the substruct (protein of interest) that contains the amide bond.
linker: The linker molecule that connects substruct to the E3 ligase.
substruct_attachment_id: The attachment point ID in the substruct substructure. E.g., 1 for the POI, as in "[*:1]".
Returns:
Tuple[Chem.Mol, Chem.Mol]: The adjusted substruct and linker molecules, in that order.
"""
# Convert dummy atoms in substruct to query atoms for substructure search
query_substruct = dummy2query(substruct)
# Identify ester group (COOR) in substruct substructure
ester_pattern = Chem.MolFromSmarts("[OX2][CX3](=[OX1])")
ester_matches = query_substruct.GetSubstructMatches(ester_pattern)
if not ester_matches:
return substruct, linker # No amide bond found, return the original substruct
side_atom = None
oxygen_idx_found, carbonyl_idx_found = None, None
for match in ester_matches:
oxygen_idx, carbonyl_idx = match[0], match[1]
oxygen_atom = query_substruct.GetAtomWithIdx(oxygen_idx)
carbonyl_atom = query_substruct.GetAtomWithIdx(carbonyl_idx)
for ester_atom in [oxygen_atom, carbonyl_atom]:
# Check neighboring atoms for attachment points
# NOTE: The dummy atom representing an attachment point have atomic number 0
for neighbor in ester_atom.GetNeighbors():
if neighbor.GetAtomicNum() == 0:
oxygen_idx_found = oxygen_idx
carbonyl_idx_found = carbonyl_idx
side_atom = "O" if ester_atom == oxygen_atom else "C"
break
# If previous search failed, check the neighbors of the neighboring
# atoms (second-order neighbors)
if oxygen_idx_found is None or carbonyl_idx_found is None:
for neighbor in ester_atom.GetNeighbors():
for second_neighbor in neighbor.GetNeighbors():
if second_neighbor.GetIdx() == carbonyl_idx or second_neighbor.GetIdx() == oxygen_idx:
continue # Skip the opposite atom from the amide bond
if second_neighbor.GetAtomicNum() == 0:
oxygen_idx_found = oxygen_idx
carbonyl_idx_found = carbonyl_idx
side_atom = "O" if ester_atom == oxygen_atom else "C"
break
else:
break
if oxygen_idx_found is None or carbonyl_idx_found is None or side_atom is None:
return substruct, linker
# Split the amide bond and adjust
dummy_label = 3
dummy_labels = [(dummy_label, dummy_label)] # The E3 and substruct will have 1 and 2, so we need a third one
amid_bond_idx = query_substruct.GetBondBetweenAtoms(oxygen_idx_found, carbonyl_idx_found).GetIdx()
fragments = Chem.FragmentOnBonds(query_substruct, [amid_bond_idx], addDummies=True, dummyLabels=dummy_labels)
# Get the fragments resulting from bond breaking
try:
mol_frags = Chem.GetMolFrags(fragments, asMols=True, sanitizeFrags=True)
except Exception as e:
if verbose:
print(e)
return substruct, linker
# Identify the "[*:substruct][<optional neighboring atom>]N[3*]" fragment, the other one will be the "truncated" substruct
ester_fragment_pattern = Chem.MolFromSmarts(f"[*:{substruct_attachment_id}][{side_atom}][{dummy_label}*]")
ester_fragment = None
substruct_fixed = None
for frag in mol_frags:
if frag.HasSubstructMatch(dummy2query(ester_fragment_pattern)):
ester_fragment = frag
else:
substruct_fixed = frag
if ester_fragment is None or substruct_fixed is None:
return substruct, linker
# In order for the function to be used "on linkers", we need to make sure
# that the ester fragment contains the attachment point of the substruct.
# If not, there's nothing to do.
if f'[*:{substruct_attachment_id}]' not in Chem.MolToSmiles(ester_fragment, canonical=True):
return substruct, linker
# Rename the "[3*]" attachment point on the amide fragment to "[*:3]"
ester_fragment_smiles = Chem.MolToSmiles(ester_fragment, canonical=True)
ester_fragment_smiles = ester_fragment_smiles.replace(f'[{dummy_label}*]', f'[*:{dummy_label}]')
ester_fragment = Chem.MolFromSmiles(ester_fragment_smiles)
# Use molzip to join the linker and the fragment at the original attachment point
linker_fixed = Chem.molzip(linker, ester_fragment)
# Rename the "[*:3]" attachment point back to the original attachment point on the linker
linker_fixed_smiles = Chem.MolToSmiles(linker_fixed, canonical=True)
linker_fixed_smiles = linker_fixed_smiles.replace(f'[*:{dummy_label}]', f'[*:{substruct_attachment_id}]')
linker_fixed = Chem.MolFromSmiles(linker_fixed_smiles)
# Rename the "[3*]" attachment point back to the original attachment point on the substruct
substruct_fixed_smiles = Chem.MolToSmiles(substruct_fixed, canonical=True)
substruct_fixed_smiles = substruct_fixed_smiles.replace(f'[{dummy_label}*]', f'[*:{substruct_attachment_id}]')
substruct_fixed = Chem.MolFromSmiles(substruct_fixed_smiles)
return substruct_fixed, linker_fixed
def adjust_ester_bonds_in_substructs(
substructs: Dict[str, str],
protac_smiles: str,
poi_attachment_id: int = 1,
e3_attachment_id: int = 2,
) -> Dict[str, str]:
""" Adjusts the ester bonds in the substructures of a PROTAC. Just a wrapper function to apply it to multiple substructures.
Args:
substructs: The substructures of the PROTAC. A dictionary of SMILES with keys 'poi', 'linker', and 'e3'.
protac_smiles: The SMILES of the PROTAC for checking reassembly.
Returns:
The updated substructures dictionary.
"""
poi_mol = Chem.MolFromSmiles(substructs['poi'])
e3_mol = Chem.MolFromSmiles(substructs['e3'])
linker_mol = Chem.MolFromSmiles(substructs['linker'])
# Fix the amide group on the POI ligand
poi_mol, linker_mol = adjust_ester_bond(poi_mol, linker_mol, poi_attachment_id)
poi_smiles = Chem.MolToSmiles(poi_mol, canonical=True)
linker_smiles = Chem.MolToSmiles(linker_mol, canonical=True)
e3_smiles = substructs['e3']
if not check_reassembly(protac_smiles, '.'.join([poi_smiles, linker_smiles, e3_smiles])):
return substructs
# Fix the amide group on the E3 binder
e3_mol, linker_mol = adjust_ester_bond(e3_mol, linker_mol, e3_attachment_id)
e3_smiles = Chem.MolToSmiles(e3_mol, canonical=True)
linker_smiles = Chem.MolToSmiles(linker_mol, canonical=True)
if not check_reassembly(protac_smiles, '.'.join([poi_smiles, linker_smiles, e3_smiles])):
return substructs
# Fix the amide group on the linker, E3 side
linker_mol, e3_mol = adjust_ester_bond(linker_mol, e3_mol, e3_attachment_id)
e3_smiles = Chem.MolToSmiles(e3_mol, canonical=True)
linker_smiles = Chem.MolToSmiles(linker_mol, canonical=True)
if not check_reassembly(protac_smiles, '.'.join([poi_smiles, linker_smiles, e3_smiles])):
return substructs
# Fix the amide group on the linker, POI side
linker_mol, poi_mol = adjust_ester_bond(linker_mol, poi_mol, poi_attachment_id)
poi_smiles = Chem.MolToSmiles(poi_mol, canonical=True)
linker_smiles = Chem.MolToSmiles(linker_mol, canonical=True)
if not check_reassembly(protac_smiles, '.'.join([poi_smiles, linker_smiles, e3_smiles])):
return substructs
substructs['poi'] = poi_smiles
substructs['e3'] = e3_smiles
substructs['linker'] = linker_smiles
return substructs |