Support allenes in canonicalizing double bonds, fixes #7044

Code/GraphMol/Canon.cpp

@@ -218,9 +218,13 @@ void canonicalizeDoubleBond(Bond *dblBond, UINT_VECT &bondVisitOrders,
   firstVisitOrder = mol.getNumBonds() + 1;
   findNeighborBonds(atom2, firstFromAtom2, secondFromAtom2, dir2Set);
 
-  // make sure we found everything we need to find:
-  CHECK_INVARIANT(firstFromAtom1, "could not find atom1");
-  CHECK_INVARIANT(firstFromAtom2, "could not find atom2");
+  // Make sure we found everything we need to find.
+  //   This really shouldn't be a problem, but molecules can end up in odd
+  //   states; for example, allenes can end up here. Instead of checking for them
+  //   explicitly, exit early in any such possible state.
+  if (!firstFromAtom1 || !firstFromAtom2) {
+    return;
+  }
 
   bool setFromBond1 = true;
   Bond::BondDir atom1Dir = Bond::NONE;

Code/GraphMol/MolStandardize/testTautomer.cpp

@@ -316,6 +316,14 @@ void testEnumerator() {
   checkAns("C/C=C/C(C)=O", {"C=C(O)C=CC", "C=CC=C(C)O", "C=CCC(=C)O",
                             "C=CCC(C)=O", "CC=CC(C)=O"});
 
+  // No stereochemistry in conjugated double bonds to nitro
+  checkAns("c1ccnc(c1)C=C[N+](=O)[O-]", {"O=[N+]([O-])C=Cc1ccccn1",
+                                         "[O-][N+](O)=C=Cc1ccccn1"});
+
+  // Retain stereochemistry in conjugated double bonds to nitro
+  checkAns("c1ccnc(c1)/C=C/[N+](=O)[O-]", {"O=[N+]([O-])/C=C/c1ccccn1",
+                                           "[O-][N+](O)=C=Cc1ccccn1"});
+
   // Remove stereochemistry from mobile double bonds
   std::string smi66 = "C/C=C\\C(C)=O";
   ROMOL_SPTR m66(SmilesToMol(smi66));
@@ -334,7 +342,7 @@ void testEnumerator() {
   std::sort(ans66.begin(), ans66.end());
   TEST_ASSERT(sm66 == ans66);
 
-  // Gaunine tautomers
+  // Guanine tautomers
   std::string smi67 = "N1C(N)=NC=2N=CNC2C1=O";
   ROMOL_SPTR m67(SmilesToMol(smi67));
   TautomerEnumeratorResult res67 = te.enumerate(*m67);
@@ -490,6 +498,56 @@ void testEnumeratorParams() {
       }
     }
   }
+  std::string eOximeSmi = "c1ccnc(c1)/C=N/O";
+  ROMOL_SPTR eOxime(SmilesToMol(eOximeSmi));
+  TEST_ASSERT(eOxime->getBondWithIdx(6)->getStereo() == Bond::STEREOE);
+  {
+    // test remove oxime E stereochemistry
+    CleanupParameters params;
+    params.tautomerRemoveBondStereo = true;
+    TautomerEnumerator te(params);
+    TautomerEnumeratorResult res = te.enumerate(*eOxime);
+    for (const auto &taut : res) {
+      TEST_ASSERT(taut->getBondWithIdx(6)->getStereo() == Bond::STEREONONE);
+    }
+  }
+  {
+    // test retain enol E stereochemistry
+    CleanupParameters params;
+    params.tautomerRemoveBondStereo = false;
+    TautomerEnumerator te(params);
+    TautomerEnumeratorResult res = te.enumerate(*eOxime);
+    for (const auto &taut : res) {
+      if (taut->getBondWithIdx(6)->getBondType() == Bond::DOUBLE) {
+        TEST_ASSERT(taut->getBondWithIdx(6)->getStereo() == Bond::STEREOE);
+      }
+    }
+  }
+  ROMOL_SPTR zOxime = "c1ccnc(c1)/C=N\\O"_smiles;
+  // zOxime->debugMol(std::cerr);
+  TEST_ASSERT(zOxime->getBondWithIdx(6)->getStereo() == Bond::STEREOZ);
+  {
+    // test remove enol Z stereochemistry
+    CleanupParameters params;
+    params.tautomerRemoveBondStereo = true;
+    TautomerEnumerator te(params);
+    TautomerEnumeratorResult res = te.enumerate(*zOxime);
+    for (const auto &taut : res) {
+      TEST_ASSERT(taut->getBondWithIdx(6)->getStereo() == Bond::STEREONONE);
+    }
+  }
+  {
+    // test retain enol Z stereochemistry
+    CleanupParameters params;
+    params.tautomerRemoveBondStereo = false;
+    TautomerEnumerator te(params);
+    TautomerEnumeratorResult res = te.enumerate(*zOxime);
+    for (const auto &taut : res) {
+      if (taut->getBondWithIdx(6)->getBondType() == Bond::DOUBLE) {
+        TEST_ASSERT(taut->getBondWithIdx(6)->getStereo() == Bond::STEREOZ);
+      }
+    }
+  }
   ROMOL_SPTR chembl2024142 =
       "[2H]C1=C(C(=C2C(=C1[2H])C(=O)C(=C(C2=O)C([2H])([2H])[2H])C/C=C(\\C)/CC([2H])([2H])/C=C(/CC/C=C(\\C)/CCC=C(C)C)\\C([2H])([2H])[2H])[2H])[2H]"_smiles;
   MolOps::RemoveHsParameters hparams;

Code/GraphMol/catch_graphmol.cpp

@@ -3475,6 +3475,19 @@ TEST_CASE(
   CHECK(bnd->getStereoAtoms().empty());
 }
 
+TEST_CASE(
+    "github #7044: canonicalization error when allenes have specified stereo") {
+  SECTION("basics") {
+    auto m = "CC=C=CC"_smiles;
+    REQUIRE(m);
+    m->getBondWithIdx(2)->setStereoAtoms(1, 4);
+    m->getBondWithIdx(2)->setStereo(Bond::STEREOCIS);
+
+    auto smi = MolToSmiles(*m);
+    CHECK(smi == "CC=C=CC");
+  }
+}
+
 TEST_CASE(
     "Github Issue #7128: ReplaceBond may cause valence issues in specific edge cases",
     "[bug][RWMol]") {