1 //
2 // Copyright (c) 2014, Novartis Institutes for BioMedical Research Inc.
3 // All rights reserved.
4 //
5 // Redistribution and use in source and binary forms, with or without
6 // modification, are permitted provided that the following conditions are
7 // met:
8 //
9 // * Redistributions of source code must retain the above copyright
10 // notice, this list of conditions and the following disclaimer.
11 // * Redistributions in binary form must reproduce the above
12 // copyright notice, this list of conditions and the following
13 // disclaimer in the documentation and/or other materials provided
14 // with the distribution.
15 // * Neither the name of Novartis Institutes for BioMedical Research Inc.
16 // nor the names of its contributors may be used to endorse or promote
17 // products derived from this software without specific prior written
18 // permission.
19 //
20 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
21 // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
22 // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
23 // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
24 // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
25 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
26 // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
27 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
28 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
29 // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
30 // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
31 //
32
33 #include <GraphMol/ChemReactions/Reaction.h>
34 #include <GraphMol/ChemReactions/ReactionUtils.h>
35 #include <GraphMol/Substruct/SubstructMatch.h>
36 #include <GraphMol/ROMol.h>
37 #include <GraphMol/Descriptors/MolDescriptors.h>
38 #include <cmath>
39
40 namespace RDKit {
41
getStartIterator(const ChemicalReaction & rxn,ReactionMoleculeType t)42 MOL_SPTR_VECT::const_iterator getStartIterator(const ChemicalReaction &rxn,
43 ReactionMoleculeType t) {
44 MOL_SPTR_VECT::const_iterator begin;
45 if (t == Reactant) {
46 begin = rxn.beginReactantTemplates();
47 }
48 if (t == Product) {
49 begin = rxn.beginProductTemplates();
50 ;
51 }
52 if (t == Agent) {
53 begin = rxn.beginAgentTemplates();
54 }
55 return begin;
56 }
57
getEndIterator(const ChemicalReaction & rxn,ReactionMoleculeType t)58 MOL_SPTR_VECT::const_iterator getEndIterator(const ChemicalReaction &rxn,
59 ReactionMoleculeType t) {
60 MOL_SPTR_VECT::const_iterator end;
61 if (t == Reactant) {
62 end = rxn.endReactantTemplates();
63 }
64 if (t == Product) {
65 end = rxn.endProductTemplates();
66 ;
67 }
68 if (t == Agent) {
69 end = rxn.endAgentTemplates();
70 }
71 return end;
72 }
73
74 namespace {
75
hasReactionMoleculeTemplateSubstructMatch(const RDKit::ChemicalReaction & rxn,const RDKit::ChemicalReaction & query_rxn,RDKit::ReactionMoleculeType t)76 bool hasReactionMoleculeTemplateSubstructMatch(
77 const RDKit::ChemicalReaction &rxn,
78 const RDKit::ChemicalReaction &query_rxn, RDKit::ReactionMoleculeType t) {
79 for (auto begin = getStartIterator(rxn, t); begin != getEndIterator(rxn, t);
80 ++begin) {
81 for (auto begin_query = getStartIterator(query_rxn, t);
82 begin_query != getEndIterator(query_rxn, t); ++begin_query) {
83 MatchVectType tvect;
84 if (SubstructMatch(*begin->get(), *begin_query->get(), tvect)) {
85 return true;
86 }
87 }
88 }
89 return false;
90 }
91 } // namespace
92
hasReactantTemplateSubstructMatch(const ChemicalReaction & rxn,const ChemicalReaction & query_rxn)93 bool hasReactantTemplateSubstructMatch(const ChemicalReaction &rxn,
94 const ChemicalReaction &query_rxn) {
95 if (rxn.getNumReactantTemplates() < query_rxn.getNumReactantTemplates()) {
96 return false;
97 }
98 if (query_rxn.getNumReactantTemplates() == 0) {
99 return true;
100 }
101 return hasReactionMoleculeTemplateSubstructMatch(rxn, query_rxn, Reactant);
102 }
103
hasProductTemplateSubstructMatch(const ChemicalReaction & rxn,const ChemicalReaction & query_rxn)104 bool hasProductTemplateSubstructMatch(const ChemicalReaction &rxn,
105 const ChemicalReaction &query_rxn) {
106 if (rxn.getNumProductTemplates() < query_rxn.getNumProductTemplates()) {
107 return false;
108 }
109 if (query_rxn.getNumProductTemplates() == 0) {
110 return true;
111 }
112 return hasReactionMoleculeTemplateSubstructMatch(rxn, query_rxn, Product);
113 }
114
hasAgentTemplateSubstructMatch(const ChemicalReaction & rxn,const ChemicalReaction & query_rxn)115 bool hasAgentTemplateSubstructMatch(const ChemicalReaction &rxn,
116 const ChemicalReaction &query_rxn) {
117 if (rxn.getNumAgentTemplates() < query_rxn.getNumAgentTemplates()) {
118 return false;
119 }
120 if (query_rxn.getNumAgentTemplates() == 0) {
121 return true;
122 }
123 return hasReactionMoleculeTemplateSubstructMatch(rxn, query_rxn, Agent);
124 }
125
hasReactionSubstructMatch(const ChemicalReaction & rxn,const ChemicalReaction & query_rxn,bool includeAgents)126 bool hasReactionSubstructMatch(const ChemicalReaction &rxn,
127 const ChemicalReaction &query_rxn,
128 bool includeAgents) {
129 if (includeAgents) {
130 return (hasReactantTemplateSubstructMatch(rxn, query_rxn) &&
131 hasProductTemplateSubstructMatch(rxn, query_rxn) &&
132 hasAgentTemplateSubstructMatch(rxn, query_rxn));
133 }
134 return (hasReactantTemplateSubstructMatch(rxn, query_rxn) &&
135 hasProductTemplateSubstructMatch(rxn, query_rxn));
136 }
137
hasReactionAtomMapping(const ChemicalReaction & rxn)138 bool hasReactionAtomMapping(const ChemicalReaction &rxn) {
139 auto begin = getStartIterator(rxn, Reactant);
140 auto end = getEndIterator(rxn, Reactant);
141 for (; begin != end; ++begin) {
142 const ROMol &reactant = *begin->get();
143 if (MolOps::getNumAtomsWithDistinctProperty(
144 reactant, common_properties::molAtomMapNumber)) {
145 return true;
146 }
147 }
148 begin = getStartIterator(rxn, Product);
149 end = getEndIterator(rxn, Product);
150 for (; begin != end; ++begin) {
151 const ROMol &reactant = *begin->get();
152 if (MolOps::getNumAtomsWithDistinctProperty(
153 reactant, common_properties::molAtomMapNumber)) {
154 return true;
155 }
156 }
157 return false;
158 }
159
isReactionTemplateMoleculeAgent(const ROMol & mol,double agentThreshold)160 bool isReactionTemplateMoleculeAgent(const ROMol &mol, double agentThreshold) {
161 unsigned numMappedAtoms = MolOps::getNumAtomsWithDistinctProperty(
162 mol, common_properties::molAtomMapNumber);
163 unsigned numAtoms = mol.getNumHeavyAtoms();
164 if (numAtoms > 0 &&
165 static_cast<double>(numMappedAtoms) / static_cast<double>(numAtoms) >=
166 agentThreshold) {
167 return false;
168 }
169 return true;
170 }
171
172 namespace {
173
getMappingNumAtomIdxMapReactants(const ChemicalReaction & rxn,std::map<int,Atom * > & reactantAtomMapping)174 void getMappingNumAtomIdxMapReactants(
175 const ChemicalReaction &rxn, std::map<int, Atom *> &reactantAtomMapping) {
176 for (auto reactIt = rxn.beginReactantTemplates();
177 reactIt != rxn.endReactantTemplates(); ++reactIt) {
178 for (const auto atom : (*reactIt)->atoms()) {
179 int reactMapNum;
180 if (atom->getPropIfPresent(common_properties::molAtomMapNumber,
181 reactMapNum)) {
182 reactantAtomMapping[reactMapNum] = atom;
183 }
184 }
185 }
186 }
187
188 // returns the atom map numbers of the neighbors of atom1 in the order in which
189 // the neighbors are attached. -1 in the vector for unmapped atoms,
190 // -1 at the end of the vector if the degree of atom1 < the degree of atom 2
getNbrOrder(const Atom * atom1,const Atom * atom2)191 std::pair<unsigned int, std::vector<int>> getNbrOrder(const Atom *atom1,
192 const Atom *atom2) {
193 std::vector<int> order;
194 order.reserve(atom1->getDegree());
195 unsigned nUnmapped = 0;
196 for (const auto &nbri : boost::make_iterator_range(
197 atom1->getOwningMol().getAtomNeighbors(atom1))) {
198 const auto &nbrAtom = atom1->getOwningMol()[nbri];
199 if (nbrAtom->getAtomMapNum() > 0) {
200 order.push_back(nbrAtom->getAtomMapNum());
201 } else {
202 order.push_back(-1);
203 ++nUnmapped;
204 }
205 }
206 if (atom1->getDegree() < atom2->getDegree()) {
207 order.push_back(-1);
208 ++nUnmapped;
209 }
210 return {nUnmapped, order};
211 }
212
checkOrderOverlap(std::vector<int> & order,unsigned int nUnmapped,const std::vector<int> & refOrder)213 bool checkOrderOverlap(std::vector<int> &order, unsigned int nUnmapped,
214 const std::vector<int> &refOrder) {
215 bool allFound = true;
216 for (auto elem : refOrder) {
217 if (elem >= 0) {
218 if (std::find(order.begin(), order.end(), elem) == order.end()) {
219 // this one was not there, is there an unmapped slot for
220 // it (i.e. a -1 value in the order)?
221 if (nUnmapped) {
222 auto negOne = std::find(order.begin(), order.end(), -1);
223 if (negOne != order.end()) {
224 *negOne = elem;
225 } else {
226 allFound = false;
227 break;
228 }
229 } else {
230 allFound = false;
231 break;
232 }
233 }
234 }
235 }
236 return allFound;
237 }
238
239 } // namespace
240
241 // returns -1 if we don't find a good match
countSwapsBetweenReactantAndProduct(const Atom * reactAtom,const Atom * prodAtom)242 int countSwapsBetweenReactantAndProduct(const Atom *reactAtom,
243 const Atom *prodAtom) {
244 PRECONDITION(reactAtom, "bad atom");
245 PRECONDITION(prodAtom, "bad atom");
246 if (reactAtom->getDegree() >= 3 && prodAtom->getDegree() >= 3 &&
247 std::abs(static_cast<int>(prodAtom->getDegree()) -
248 static_cast<int>(reactAtom->getDegree())) <= 1) {
249 std::vector<int> reactOrder;
250 unsigned int nReactUnmapped;
251 std::tie(nReactUnmapped, reactOrder) = getNbrOrder(reactAtom, prodAtom);
252 if (nReactUnmapped <= 1) {
253 std::vector<int> prodOrder;
254 unsigned int nProdUnmapped;
255 std::tie(nProdUnmapped, prodOrder) = getNbrOrder(prodAtom, reactAtom);
256 if (nProdUnmapped <= 1) {
257 // check that each element of the product mappings is
258 // in the reactant mappings
259 if (checkOrderOverlap(reactOrder, nReactUnmapped, prodOrder)) {
260 // found a match for all the product atoms, what about all
261 // the reactant atoms?
262 if (checkOrderOverlap(prodOrder, nProdUnmapped, reactOrder)) {
263 return countSwapsToInterconvert(reactOrder, prodOrder);
264 }
265 }
266 }
267 }
268 }
269 return -1;
270 }
271
updateProductsStereochem(ChemicalReaction * rxn)272 void updateProductsStereochem(ChemicalReaction *rxn) {
273 std::map<int, Atom *> reactantMapping;
274 getMappingNumAtomIdxMapReactants(*rxn, reactantMapping);
275 for (MOL_SPTR_VECT::const_iterator prodIt = rxn->beginProductTemplates();
276 prodIt != rxn->endProductTemplates(); ++prodIt) {
277 for (auto prodAtom : (*prodIt)->atoms()) {
278 if (prodAtom->hasProp(common_properties::molInversionFlag)) {
279 continue;
280 }
281 if (!prodAtom->hasProp(common_properties::molAtomMapNumber)) {
282 // if we have stereochemistry specified, it's automatically
283 // creating stereochem:
284 prodAtom->setProp(common_properties::molInversionFlag, 4);
285 continue;
286 }
287 int mapNum;
288 prodAtom->getProp(common_properties::molAtomMapNumber, mapNum);
289 if (reactantMapping.find(mapNum) != reactantMapping.end()) {
290 const auto reactAtom = reactantMapping[mapNum];
291 if (prodAtom->getChiralTag() != Atom::CHI_UNSPECIFIED &&
292 prodAtom->getChiralTag() != Atom::CHI_OTHER) {
293 if (reactAtom->getChiralTag() != Atom::CHI_UNSPECIFIED &&
294 reactAtom->getChiralTag() != Atom::CHI_OTHER) {
295 // both have stereochem specified, we're either preserving
296 // or inverting
297 if (reactAtom->getChiralTag() == prodAtom->getChiralTag()) {
298 prodAtom->setProp(common_properties::molInversionFlag, 2);
299 } else {
300 // FIX: this is technically fragile: it should be checking
301 // if the atoms both have tetrahedral chirality. However,
302 // at the moment that's the only chirality available, so
303 // there's no need to go monkeying around.
304 prodAtom->setProp(common_properties::molInversionFlag, 1);
305 }
306
307 // FIX this should move out into a separate function
308 // last thing to check here: if the ordering of the bonds
309 // around the atom changed from reactants->products then we
310 // may need to adjust the inversion flag
311 int nSwaps =
312 countSwapsBetweenReactantAndProduct(reactAtom, prodAtom);
313 if (nSwaps >= 0 && nSwaps % 2) {
314 auto mival =
315 prodAtom->getProp<int>(common_properties::molInversionFlag);
316 if (mival == 1) {
317 prodAtom->setProp(common_properties::molInversionFlag, 2);
318 } else if (mival == 2) {
319 prodAtom->setProp(common_properties::molInversionFlag, 1);
320 } else {
321 CHECK_INVARIANT(false, "inconsistent molInversionFlag");
322 }
323 }
324 } else {
325 // stereochem in the product, but not in the reactant
326 prodAtom->setProp(common_properties::molInversionFlag, 4);
327 }
328 } else if (reactantMapping[mapNum]->getChiralTag() !=
329 Atom::CHI_UNSPECIFIED &&
330 reactantMapping[mapNum]->getChiralTag() != Atom::CHI_OTHER) {
331 // stereochem in the reactant, but not the product:
332 prodAtom->setProp(common_properties::molInversionFlag, 3);
333 }
334 } else {
335 // introduction of new stereocenter by the reaction
336 prodAtom->setProp(common_properties::molInversionFlag, 4);
337 }
338 }
339 }
340 }
341
342 namespace {
343
removeMappingNumbersFromReactionMoleculeTemplate(const MOL_SPTR_VECT & molVec)344 void removeMappingNumbersFromReactionMoleculeTemplate(
345 const MOL_SPTR_VECT &molVec) {
346 for (const auto &begin : molVec) {
347 ROMol &mol = *begin.get();
348 for (ROMol::AtomIterator atomIt = mol.beginAtoms();
349 atomIt != mol.endAtoms(); ++atomIt) {
350 if ((*atomIt)->hasProp(common_properties::molAtomMapNumber)) {
351 (*atomIt)->clearProp(common_properties::molAtomMapNumber);
352 }
353 }
354 }
355 }
356
357 } // namespace
358
removeMappingNumbersFromReactions(const ChemicalReaction & rxn)359 void removeMappingNumbersFromReactions(const ChemicalReaction &rxn) {
360 removeMappingNumbersFromReactionMoleculeTemplate(rxn.getAgents());
361 removeMappingNumbersFromReactionMoleculeTemplate(rxn.getProducts());
362 removeMappingNumbersFromReactionMoleculeTemplate(rxn.getReactants());
363 }
364
365 } // namespace RDKit
366