double AlignMolecule(ROMol &prbMol, const ROMol &refMol, int prbCid = -1,
int refCid = -1, python::object atomMap = python::list(),
python::object weights = python::list(),
bool reflect = false, unsigned int maxIters = 50) {
MatchVectType *aMap = _translateAtomMap(atomMap);
unsigned int nAtms;
if (aMap) {
nAtms = aMap->size();
} else {
nAtms = prbMol.getNumAtoms();
}
RDNumeric::DoubleVector *wtsVec = _translateWeights(weights);
if (wtsVec) {
if (wtsVec->size() != nAtms) {
throw_value_error("Incorrect number of weights specified");
}
}
double rmsd;
{
NOGIL gil;
rmsd = MolAlign::alignMol(prbMol, refMol, prbCid, refCid, aMap, wtsVec,
reflect, maxIters);
}
if (aMap) {
delete aMap;
}
if (wtsVec) {
delete wtsVec;
}
return rmsd;
}
void testMMFFO3AConstraintsAndLocalOnly() {
std::string rdbase = getenv("RDBASE");
std::string sdf = rdbase + "/Code/GraphMol/MolAlign/test_data/ref_e2.sdf";
SDMolSupplier supplier(sdf, true, false);
const int refNum = 23;
const int prbNum = 32;
ROMol *refMol = supplier[refNum];
ROMol *prbMol = supplier[prbNum];
unsigned int refNAtoms = refMol->getNumAtoms();
std::vector<double> refLogpContribs(refNAtoms);
std::vector<double> refMRContribs(refNAtoms);
std::vector<unsigned int> refAtomTypes(refNAtoms);
std::vector<std::string> refAtomTypeLabels(refNAtoms);
Descriptors::getCrippenAtomContribs(*refMol, refLogpContribs,
refMRContribs, true, &refAtomTypes, &refAtomTypeLabels);
unsigned int prbNAtoms = prbMol->getNumAtoms();
std::vector<double> prbLogpContribs(prbNAtoms);
std::vector<double> prbMRContribs(prbNAtoms);
std::vector<unsigned int> prbAtomTypes(prbNAtoms);
std::vector<std::string> prbAtomTypeLabels(prbNAtoms);
Descriptors::getCrippenAtomContribs(*prbMol, prbLogpContribs,
prbMRContribs, true, &prbAtomTypes, &prbAtomTypeLabels);
RWMol *patt = SmartsToMol("S");
MatchVectType matchVect;
TEST_ASSERT(SubstructMatch(*refMol, (ROMol &)*patt, matchVect));
delete patt;
unsigned int refSIdx = matchVect[0].second;
matchVect.clear();
patt = SmartsToMol("O");
TEST_ASSERT(SubstructMatch(*prbMol, (ROMol &)*patt, matchVect));
delete patt;
unsigned int prbOIdx = matchVect[0].second;
std::vector<double> distOS(2);
distOS[0] = 2.7;
distOS[1] = 0.4;
std::vector<double> weights(2);
weights[0] = 0.1;
weights[1] = 100.0;
for (unsigned int i = 0; i < 2; ++i) {
MatchVectType constraintMap;
constraintMap.push_back(std::make_pair(prbOIdx, refSIdx));
RDNumeric::DoubleVector constraintWeights(1);
constraintWeights[0] = weights[i];
MolAlign::O3A *o3a = new MolAlign::O3A(*prbMol, *refMol,
&prbLogpContribs, &refLogpContribs, MolAlign::O3A::CRIPPEN,
-1, -1, false, 50, 0, &constraintMap, &constraintWeights);
TEST_ASSERT(o3a);
o3a->align();
delete o3a;
o3a = new MolAlign::O3A(*prbMol, *refMol, &prbLogpContribs,
&refLogpContribs, MolAlign::O3A::CRIPPEN, -1, -1,
false, 50, MolAlign::O3_LOCAL_ONLY);
TEST_ASSERT(o3a);
o3a->align();
delete o3a;
double d = (prbMol->getConformer().getAtomPos(prbOIdx)
- refMol->getConformer().getAtomPos(refSIdx)).length();
TEST_ASSERT(feq(d, distOS[i], 0.1));
}
}
void testDativeMatch() {
BOOST_LOG(rdErrorLog) << "-------------------------------------" << std::endl;
BOOST_LOG(rdErrorLog) << " Test dative-bond matching" << std::endl;
{
std::string smi = "[Cu]->[Fe]";
ROMol *mol = SmilesToMol(smi);
TEST_ASSERT(mol);
// make sure a self-match works
MatchVectType match;
TEST_ASSERT(SubstructMatch(*mol, *mol, match));
TEST_ASSERT(match.size() == mol->getNumAtoms());
{ // reverse the order and make sure that works
std::string sma = "[Fe]<-[Cu]";
ROMol *qmol = SmilesToMol(sma);
TEST_ASSERT(qmol);
MatchVectType match;
TEST_ASSERT(SubstructMatch(*mol, *qmol, match));
TEST_ASSERT(match.size() == qmol->getNumAtoms());
delete qmol;
}
{ // reverse the direction and make sure that does not work.
std::string sma = "[Fe]->[Cu]";
ROMol *qmol = SmilesToMol(sma);
TEST_ASSERT(qmol);
MatchVectType match;
TEST_ASSERT(!SubstructMatch(*mol, *qmol, match));
delete qmol;
}
delete mol;
}
BOOST_LOG(rdErrorLog) << " done" << std::endl;
}
PyObject *getMolAlignTransform(const ROMol &prbMol, const ROMol &refMol,
int prbCid = -1, int refCid = -1,
python::object atomMap = python::list(),
python::object weights = python::list(),
bool reflect = false,
unsigned int maxIters = 50) {
MatchVectType *aMap = _translateAtomMap(atomMap);
unsigned int nAtms;
if (aMap) {
nAtms = aMap->size();
} else {
nAtms = prbMol.getNumAtoms();
}
RDNumeric::DoubleVector *wtsVec = _translateWeights(weights);
if (wtsVec) {
if (wtsVec->size() != nAtms) {
throw_value_error("Incorrect number of weights specified");
}
}
RDGeom::Transform3D trans;
double rmsd;
{
NOGIL gil;
rmsd = MolAlign::getAlignmentTransform(
prbMol, refMol, trans, prbCid, refCid, aMap, wtsVec, reflect, maxIters);
}
if (aMap) {
delete aMap;
}
if (wtsVec) {
delete wtsVec;
}
return generateRmsdTransPyTuple(rmsd, trans);
}
void test4() {
std::cout << " ----------------- Test 4" << std::endl;
MatchVectType matchV;
std::vector<MatchVectType> matches;
int n;
bool updateLabel = true;
bool takeOwnership = true;
RWMol *m, *q1, *q2;
auto *a6 = new Atom(6);
auto *a8 = new Atom(8);
m = new RWMol();
m->addAtom(a6);
m->addAtom(a6);
m->addAtom(a8);
m->addAtom(a6);
m->addAtom(a6);
m->addBond(1, 0, Bond::SINGLE);
m->addBond(1, 2, Bond::SINGLE);
m->addBond(1, 3, Bond::SINGLE);
m->addBond(2, 4, Bond::SINGLE);
// this will be the recursive query
q1 = new RWMol();
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addAtom(new QueryAtom(8), updateLabel, takeOwnership);
q1->addBond(0, 1, Bond::UNSPECIFIED);
// here's the main query
q2 = new RWMol();
auto *qA = new QueryAtom(6);
auto *rsq = new RecursiveStructureQuery(q1);
qA->expandQuery(rsq, Queries::COMPOSITE_AND);
// std::cout << "post expand: " << qA->getQuery() << std::endl;
q2->addAtom(qA, true, true);
// std::cout << "mol: " << q2->getAtomWithIdx(0)->getQuery() << std::endl;
q2->addAtom(new QueryAtom(6), true, true);
q2->addBond(0, 1, Bond::UNSPECIFIED);
bool found = SubstructMatch(*m, *q2, matchV);
CHECK_INVARIANT(found, "");
CHECK_INVARIANT(matchV.size() == 2, "");
TEST_ASSERT(matchV[0].first == 0);
TEST_ASSERT(matchV[0].second == 1);
TEST_ASSERT(matchV[1].first == 1);
TEST_ASSERT(matchV[1].second == 0 || matchV[1].second == 3);
n = SubstructMatch(*m, *q2, matches, true);
TEST_ASSERT(n == 2);
TEST_ASSERT(matches.size() == (size_t)n);
TEST_ASSERT(matches[0].size() == 2);
TEST_ASSERT(matches[1].size() == 2);
TEST_ASSERT(matches[0][0].second == matches[1][0].second);
TEST_ASSERT(matches[0][1].second != matches[1][1].second);
delete m;
delete a6;
delete a8;
delete q2;
std::cout << "Done\n" << std::endl;
}
void testSubstructMatchDMAP() {
BOOST_LOG(rdInfoLog) << "-----------------------\n"
<< "testSubstructMatchDMAP" << std::endl;
RWMol *mol = SmilesToMol("C(C)Nc1cc[nH+]cc1");
RWMol *query = SmartsToMol("[#7+]");
ResonanceMolSupplier *resMolSuppl;
unsigned int n;
MatchVectType p;
resMolSuppl = new ResonanceMolSupplier((ROMol &)*mol);
std::vector<MatchVectType> matchVect;
n = SubstructMatch(*mol, *query, matchVect, false, true, false, false);
TEST_ASSERT((n == 1) && (matchVect.size() == 1));
p = matchVect[0];
TEST_ASSERT((p.size() == 1) && (p[0].second == 6));
matchVect.clear();
n = SubstructMatch(*resMolSuppl, *query, matchVect, false, true, false,
false);
TEST_ASSERT((n == 2) && (matchVect.size() == 2));
std::vector<unsigned int> v;
p = matchVect[0];
TEST_ASSERT(p.size() == 1);
v.push_back(p[0].second);
p = matchVect[1];
TEST_ASSERT(p.size() == 1);
v.push_back(p[0].second);
std::sort(v.begin(), v.end());
TEST_ASSERT(v[0] == 2);
TEST_ASSERT(v[1] == 6);
delete mol;
delete query;
delete resMolSuppl;
}
double getAlignmentTransform(const ROMol &prbMol, const ROMol &refMol,
RDGeom::Transform3D &trans,
int prbCid, int refCid,
const MatchVectType *atomMap,
const RDNumeric::DoubleVector *weights, bool reflect,
unsigned int maxIterations) {
RDGeom::Point3DConstPtrVect refPoints, prbPoints;
const Conformer &prbCnf = prbMol.getConformer(prbCid);
const Conformer &refCnf = refMol.getConformer(refCid);
if (atomMap == 0) {
// we have to figure out the mapping between the two molecule
MatchVectType match;
if (SubstructMatch(refMol, prbMol, match)) {
MatchVectType::const_iterator mi;
for (mi = match.begin(); mi != match.end(); mi++) {
prbPoints.push_back(&prbCnf.getAtomPos(mi->first));
refPoints.push_back(&refCnf.getAtomPos(mi->second));
}
} else {
throw MolAlignException("No sub-structure match found between the probe and query mol");
}
} else {
MatchVectType::const_iterator mi;
for (mi = atomMap->begin(); mi != atomMap->end(); mi++) {
prbPoints.push_back(&prbCnf.getAtomPos(mi->first));
refPoints.push_back(&refCnf.getAtomPos(mi->second));
}
}
double ssr = RDNumeric::Alignments::AlignPoints(refPoints, prbPoints,
trans, weights, reflect, maxIterations);
ssr /= (prbPoints.size());
return sqrt(ssr);
}
ROMol *prepareMol(const ROMol &mol, const FragCatParams *fparams,
MatchVectType &aToFmap) {
PRECONDITION(fparams,"");
// get a mapping of the functional groups onto the molecule
INT_VECT fgBonds;
MatchVectType aidToFid = findFuncGroupsOnMol(mol, fparams, fgBonds);
// get the core piece of molecule (i.e. the part of the molecule
// without the functional groups). This basically the part of the molecule
// that does not contain the function group bonds given by "fgBonds"
INT_VECT cBonds;
int bid, nbds = mol.getNumBonds();
for (bid = 0; bid < nbds; bid++) {
if (std::find(fgBonds.begin(), fgBonds.end(), bid) == fgBonds.end()) {
cBonds.push_back(bid);
}
}
INT_MAP_INT aIdxMap; // a map from atom id in mol to the new atoms id in coreMol
ROMol *coreMol = Subgraphs::pathToSubmol(mol, cBonds, false, aIdxMap);
// now map the functional groups on mol to coreMol using aIdxMap
MatchVectType::iterator mati;
int newID;
for (mati = aidToFid.begin(); mati != aidToFid.end(); mati++) {
newID = aIdxMap[mati->first];
aToFmap.push_back(std::pair<int, int>(newID, mati->second));
}
return coreMol;
}
//*************************************************************************************
//
// Adds 2D coordinates to a molecule using the Depict.dll
//
// ARGUMENTS:
// mol: the molecule to be altered
// tempFilename: (OPTIONAL) the name of the temporary file
//
// RETURNS:
// 1 on success, 0 otherwise
//
// Here's the process by which this works (it's kind of contorted):
// 1) convert the mol to SMILES
// 2) use the DLL to convert the SMILES to a mol file (on disk)
// 3) parse the mol file into a temporary molecule
// 4) do a substructure match from the old molecule to the
// temp one (which may have a different atom numbering or additional
// atoms added).
// 5) Update the positions of the atoms on the old molecule.
// 6) Free the temp molecule.
//
// NOTES:
// - *FIX:* at the moment we're not doing anything to clear up the
// temp file created in this process. It'll always have the same
// name unless the user explicitly asks that we do something different.
// - To use the DLL, it's essential that the COMBICHEM_ROOT and COMBICHEM_RELEASE
// environment variables be set. If this isn't done, this whole process
// will fail.
// - See the notes above about failures when opening the DLL.
//
//*************************************************************************************
int Add2DCoordsToMolDLL(ROMol &mol,std::string tempFilename){
std::string smi=MolToSmiles(mol,true);
int tmp = SmilesToMolFileDLL(smi,tempFilename);
int res = -1;
if(tmp){
// build another mol from that mol file:
RWMol *tmpMol = MolFileToMol(tempFilename,false);
// match it up with the starting mol:
// (We need to do this because the depict.dll conversion
// to a mol file may have added Hs)
MatchVectType matchVect;
bool hasMatch=SubstructMatch(tmpMol,&mol,matchVect);
if(hasMatch){
const Conformer &conf = tmpMol->getCoformer(0);
Coformer *nconf = new Coformer(mol.getNumAtoms());
for(MatchVectType::const_iterator mvi=matchVect.begin();
mvi!=matchVect.end();mvi++){
nconf->setAtomPos(conf.getAtomPos(mvi->first));
}
confId = (int)mol.addConformer(nconf, true);
}
delete tmpMol;
}
return res;
}
bool StructCheckTautomer::applyTautomer(unsigned it) {
if (Options.FromTautomer.size() <= it || Options.ToTautomer.size() <= it) {
if (Options.Verbose)
BOOST_LOG(rdInfoLog) << "ERROR: incorrect Tautomer index it=" << it
<< "\n";
return false;
}
const ROMol &fromTautomer = *Options.FromTautomer[it];
const ROMol &toTautomer = *Options.ToTautomer[it];
if (toTautomer.getNumAtoms() != fromTautomer.getNumAtoms()) {
if (Options.Verbose)
BOOST_LOG(rdInfoLog) << "ERROR: incorrect data toTautomer.getNumAtoms() "
"!= fromTautomer.getNumAtoms()\n";
// incorrect data
// throw(.....);
return false;
}
const unsigned nta = toTautomer.getNumAtoms();
const unsigned ntb = toTautomer.getNumBonds();
MatchVectType match; // The format is (queryAtomIdx, molAtomIdx)
std::vector<unsigned> atomIdxMap(
Mol.getNumAtoms()); // matched tau atom indeces
if (!SubstructMatch(Mol, *Options.FromTautomer[it],
match)) // SSMatch(mp, from_tautomer, SINGLE_MATCH);
return false;
if (Options.Verbose)
BOOST_LOG(rdInfoLog) << "found match for from_tautomer with " << nta
<< " atoms\n";
// init
for (unsigned i = 0; i < Mol.getNumAtoms(); i++) atomIdxMap[i] = -1;
for (MatchVectType::const_iterator mit = match.begin(); mit != match.end();
mit++) {
unsigned tai = mit->first; // From and To Tautomer Atom index
unsigned mai = mit->second; // Mol Atom index
atomIdxMap[mai] = tai;
}
// scan for completely mapped bonds and replace bond order with mapped bond
// from to_tautomer
for (RDKit::BondIterator_ bond = Mol.beginBonds(); bond != Mol.endBonds();
bond++) {
unsigned ti = atomIdxMap[(*bond)->getBeginAtomIdx()];
unsigned tj = atomIdxMap[(*bond)->getEndAtomIdx()];
if (-1 == ti || -1 == tj) continue;
const Bond *tb = toTautomer.getBondBetweenAtoms(ti, tj);
if (tb && (*bond)->getBondType() != tb->getBondType()) {
(*bond)->setBondType(tb->getBondType());
}
}
// apply charge/radical fixes if any
for (unsigned i = 0; i < match.size(); i++) {
Atom &atom = *Mol.getAtomWithIdx(match[i].second);
const Atom &ta = *toTautomer.getAtomWithIdx(match[i].first);
atom.setFormalCharge(ta.getFormalCharge());
atom.setNumRadicalElectrons(ta.getNumRadicalElectrons());
}
return true;
}
void test2(){
std::cout << " ----------------- Test 2" << std::endl;
MatchVectType matchV;
std::vector< MatchVectType > matches;
unsigned int n;
RWMol *m,*q1;
m = new RWMol();
m->addAtom(new Atom(6));
m->addAtom(new Atom(6));
m->addAtom(new Atom(8));
m->addBond(0,1,Bond::SINGLE);
m->addBond(1,2,Bond::SINGLE);
q1 = new RWMol();
q1->addAtom(new QueryAtom(6));
q1->addAtom(new QueryAtom(8));
q1->addBond(0,1,Bond::SINGLE);
n = SubstructMatch(*m,*q1,matchV);
TEST_ASSERT(n);
TEST_ASSERT(matchV.size()==2);
TEST_ASSERT(matchV[0].first==0);
TEST_ASSERT(matchV[0].second==1);
TEST_ASSERT(matchV[1].first==1);
TEST_ASSERT(matchV[1].second==2);
n = SubstructMatch(*m,*q1,matches,false);
CHECK_INVARIANT(n==1,"");
CHECK_INVARIANT(matches.size()==n,"");
CHECK_INVARIANT(matches[0].size()==2,"");
n = SubstructMatch(*m,*q1,matches,true);
CHECK_INVARIANT(n==1,"");
CHECK_INVARIANT(matches.size()==n,"");
CHECK_INVARIANT(matches[0].size()==2,"");
CHECK_INVARIANT(SubstructMatch(*m,*q1,matchV),"");
CHECK_INVARIANT(matchV.size()==2,"");
delete m;
m = new RWMol();
m->addAtom(new Atom(6));
m->addAtom(new Atom(6));
m->addAtom(new Atom(8));
m->addBond(0,1,Bond::SINGLE);
m->addBond(1,2,Bond::DOUBLE);
matches.clear();
n = SubstructMatch(*m,*q1,matches,false);
CHECK_INVARIANT(n==0,"");
CHECK_INVARIANT(matches.size()==n,"");
n = SubstructMatch(*m,*q1,matches,true);
CHECK_INVARIANT(n==0,"");
CHECK_INVARIANT(matches.size()==n,"");
CHECK_INVARIANT(!SubstructMatch(*m,*q1,matchV),"");
std::cout << "Done\n" << std::endl;
}
void testRecursiveSerialNumbers(){
std::cout << " ----------------- Testing serial numbers on recursive queries" << std::endl;
MatchVectType matchV;
std::vector< MatchVectType > matches;
int n;
RWMol *m,*q1,*q2;
Atom *a6 = new Atom(6);
Atom *a8 = new Atom(8);
m = new RWMol();
m->addAtom(a6);
m->addAtom(a6);
m->addAtom(a8);
m->addAtom(a6);
m->addAtom(a6);
m->addBond(1,0,Bond::SINGLE);
m->addBond(1,2,Bond::SINGLE);
m->addBond(1,3,Bond::SINGLE);
m->addBond(2,4,Bond::SINGLE);
{
// this will be the recursive query
q1 = new RWMol();
q1->addAtom(new QueryAtom(6),true);
q1->addAtom(new QueryAtom(8),true);
q1->addBond(0,1,Bond::UNSPECIFIED);
// here's the main query
q2 = new RWMol();
QueryAtom *qA = new QueryAtom(6);
RecursiveStructureQuery *rsq = new RecursiveStructureQuery(new RWMol(*q1),1);
qA->expandQuery(rsq,Queries::COMPOSITE_AND);
//std::cout << "post expand: " << qA->getQuery() << std::endl;
q2->addAtom(qA,true,true);
//std::cout << "mol: " << q2->getAtomWithIdx(0)->getQuery() << std::endl;
q2->addAtom(new QueryAtom(8),true,true);
q2->addBond(0,1,Bond::UNSPECIFIED);
qA = new QueryAtom(6);
rsq = new RecursiveStructureQuery(new RWMol(*q1),1);
qA->expandQuery(rsq,Queries::COMPOSITE_AND);
q2->addAtom(qA,true,true);
q2->addBond(1,2,Bond::UNSPECIFIED);
bool found = SubstructMatch(*m,*q2,matchV);
CHECK_INVARIANT(found,"");
CHECK_INVARIANT(matchV.size()==3,"");
n = SubstructMatch(*m,*q2,matches,true);
TEST_ASSERT(n==1);
TEST_ASSERT(matches.size()==1);
TEST_ASSERT(matches[0].size()==3);
delete q1;
delete q2;
}
delete m;
std::cout << "Done\n" << std::endl;
}
void test4() {
BOOST_LOG(rdErrorLog) << "-------------------------------------" << std::endl;
BOOST_LOG(rdErrorLog) << "FeatureDef functionality testing." << std::endl;
ROMol *testMol;
MatchVectType mv;
std::string inText;
int res;
MolChemicalFeatureDef::CollectionType featureDefs;
MolChemicalFeatureDef::CollectionType::const_iterator featDefIt;
MolChemicalFeatureDef::CollectionType::value_type featDef;
inText =
"DefineFeature HDonor1 [N,O;!H0]\n"
" Family HBondDonor\n"
" Weights 1.0\n"
"EndFeature\n"
"DefineFeature HAcceptor1 [N,O]\n"
" Family HAcceptor\n"
" Weights 1.0\n"
"EndFeature\n";
res = parseFeatureData(inText, featureDefs);
TEST_ASSERT(!res);
TEST_ASSERT(featureDefs.size() == 2);
featDefIt = featureDefs.begin();
featDef = *featDefIt;
TEST_ASSERT(featDef->getSmarts() == "[N,O;!H0]");
TEST_ASSERT(featDef->getPattern());
TEST_ASSERT(featDef->getPattern()->getNumAtoms() == 1);
TEST_ASSERT(featDef->getNumWeights() == 1);
featDef = *(++featureDefs.begin());
TEST_ASSERT(featDef->getSmarts() == "[N,O]");
TEST_ASSERT(featDef->getPattern());
TEST_ASSERT(featDef->getPattern()->getNumAtoms() == 1);
TEST_ASSERT(featDef->getNumWeights() == 1);
testMol = SmilesToMol("COCN");
TEST_ASSERT(testMol);
featDef = *featureDefs.begin();
TEST_ASSERT(SubstructMatch(*testMol, *featDef->getPattern(), mv));
TEST_ASSERT(mv.size() == 1);
TEST_ASSERT(mv[0].first == 0);
TEST_ASSERT(mv[0].second == 3);
featDef = *(++featureDefs.begin());
TEST_ASSERT(SubstructMatch(*testMol, *featDef->getPattern(), mv));
TEST_ASSERT(mv.size() == 1);
TEST_ASSERT(mv[0].first == 0);
TEST_ASSERT(mv[0].second == 1 || mv[0].second == 3);
delete testMol;
BOOST_LOG(rdErrorLog) << " done" << std::endl;
}
void test5QueryRoot() {
std::cout << " ----------------- Test 5 QueryRoot" << std::endl;
MatchVectType matchV;
std::vector<MatchVectType> matches;
int n;
bool updateLabel = true;
bool takeOwnership = true;
RWMol *m, *q1, *q2;
auto *a6 = new Atom(6);
auto *a8 = new Atom(8);
// CC(OC)C
m = new RWMol();
m->addAtom(a6);
m->addAtom(a6);
m->addAtom(a8);
m->addAtom(a6);
m->addAtom(a6);
m->addBond(0, 1, Bond::SINGLE);
m->addBond(1, 2, Bond::SINGLE);
m->addBond(1, 4, Bond::SINGLE);
m->addBond(2, 3, Bond::SINGLE);
// this will be the recursive query
q1 = new RWMol();
q1->addAtom(new QueryAtom(8), updateLabel, takeOwnership);
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addBond(0, 1, Bond::UNSPECIFIED);
q1->setProp(common_properties::_queryRootAtom, 1);
// here's the main query
q2 = new RWMol();
auto *qA = new QueryAtom();
auto *rsq = new RecursiveStructureQuery(q1);
qA->setQuery(rsq);
q2->addAtom(qA, true, true);
q2->addAtom(new QueryAtom(6), true, true);
q2->addBond(0, 1, Bond::UNSPECIFIED);
bool found = SubstructMatch(*m, *q2, matchV);
CHECK_INVARIANT(found, "");
CHECK_INVARIANT(matchV.size() == 2, "");
n = SubstructMatch(*m, *q2, matches, true);
CHECK_INVARIANT(n == 2, "");
CHECK_INVARIANT(matches[0].size() == 2, "");
delete m;
delete a6;
delete a8;
delete q2;
std::cout << "Done\n" << std::endl;
}
void testCrippenO3AConstraints() {
ROMol *m = SmilesToMol("n1ccc(cc1)-c1ccccc1");
TEST_ASSERT(m);
ROMol *m1 = MolOps::addHs(*m);
delete m;
TEST_ASSERT(m1);
DGeomHelpers::EmbedMolecule(*m1);
MMFF::sanitizeMMFFMol((RWMol &)(*m1));
MMFF::MMFFMolProperties mp(*m1);
TEST_ASSERT(mp.isValid());
ForceFields::ForceField *field = MMFF::constructForceField(*m1, &mp);
field->initialize();
field->minimize();
RWMol *patt = SmartsToMol("nccc-cccc");
MatchVectType matchVect;
TEST_ASSERT(SubstructMatch(*m1, (ROMol &)*patt, matchVect));
unsigned int nIdx = matchVect[0].second;
unsigned int cIdx = matchVect[matchVect.size() - 1].second;
MolTransforms::setDihedralDeg(m1->getConformer(), matchVect[2].second,
matchVect[3].second, matchVect[4].second, matchVect[5].second, 0.0);
ROMol m2(*m1);
MolAlign::randomTransform(m2);
ROMol m3(m2);
unsigned int prbNAtoms = m2.getNumAtoms();
std::vector<double> prbLogpContribs(prbNAtoms);
std::vector<double> prbMRContribs(prbNAtoms);
std::vector<unsigned int> prbAtomTypes(prbNAtoms);
std::vector<std::string> prbAtomTypeLabels(prbNAtoms);
Descriptors::getCrippenAtomContribs(m2, prbLogpContribs,
prbMRContribs, true, &prbAtomTypes, &prbAtomTypeLabels);
MolAlign::O3A *o3a = new MolAlign::O3A(m2, *m1,
&prbLogpContribs, &prbLogpContribs, MolAlign::O3A::CRIPPEN);
TEST_ASSERT(o3a);
o3a->align();
delete o3a;
double d = (m2.getConformer().getAtomPos(cIdx)
- m1->getConformer().getAtomPos(cIdx)).length();
TEST_ASSERT(feq(d, 0.0, 1));
MatchVectType constraintMap;
constraintMap.push_back(std::make_pair(cIdx, nIdx));
o3a = new MolAlign::O3A(m3, *m1, &prbLogpContribs, &prbLogpContribs,
MolAlign::O3A::CRIPPEN, -1, -1, false, 50, 0, &constraintMap);
TEST_ASSERT(o3a);
o3a->align();
delete o3a;
d = (m3.getConformer().getAtomPos(cIdx)
- m1->getConformer().getAtomPos(cIdx)).length();
TEST_ASSERT(feq(d, 7.0, 1.0));
delete m1;
}
void testCrippenO3AConstraintsAndLocalOnly() {
std::string rdbase = getenv("RDBASE");
std::string sdf = rdbase + "/Code/GraphMol/MolAlign/test_data/ref_e2.sdf";
SDMolSupplier supplier(sdf, true, false);
const int refNum = 23;
const int prbNum = 32;
ROMol *refMol = supplier[refNum];
ROMol *prbMol = supplier[prbNum];
MMFF::MMFFMolProperties refMP(*refMol);
TEST_ASSERT(refMP.isValid());
MMFF::MMFFMolProperties prbMP(*prbMol);
TEST_ASSERT(prbMP.isValid());
RWMol *patt = SmartsToMol("S");
MatchVectType matchVect;
TEST_ASSERT(SubstructMatch(*refMol, (ROMol &)*patt, matchVect));
delete patt;
unsigned int refSIdx = matchVect[0].second;
matchVect.clear();
patt = SmartsToMol("O");
TEST_ASSERT(SubstructMatch(*prbMol, (ROMol &)*patt, matchVect));
delete patt;
unsigned int prbOIdx = matchVect[0].second;
std::vector<double> distOS(2);
distOS[0] = 3.2;
distOS[1] = 0.3;
std::vector<double> weights(2);
weights[0] = 10.0;
weights[1] = 100.0;
for (unsigned int i = 0; i < 2; ++i) {
MatchVectType constraintMap;
constraintMap.push_back(std::make_pair(prbOIdx, refSIdx));
RDNumeric::DoubleVector constraintWeights(1);
constraintWeights[0] = weights[i];
MolAlign::O3A *o3a = new MolAlign::O3A(*prbMol, *refMol, &prbMP, &refMP,
MolAlign::O3A::MMFF94, -1, -1, false, 50, 0, &constraintMap, &constraintWeights);
TEST_ASSERT(o3a);
o3a->align();
delete o3a;
o3a = new MolAlign::O3A(*prbMol, *refMol, &prbMP, &refMP,
MolAlign::O3A::MMFF94, -1, -1, false, 50, MolAlign::O3_LOCAL_ONLY);
TEST_ASSERT(o3a);
o3a->align();
delete o3a;
double d = (prbMol->getConformer().getAtomPos(prbOIdx)
- refMol->getConformer().getAtomPos(refSIdx)).length();
TEST_ASSERT(feq(d, distOS[i], 0.1));
}
}
void test6() {
std::cout << " ----------------- Test 6 (Issue71 related)" << std::endl;
MatchVectType matchV;
std::vector<MatchVectType> matches;
int n;
bool updateLabel = true;
bool takeOwnership = true;
RWMol *m, *q1;
auto *a6 = new Atom(6);
m = new RWMol();
m->addAtom(a6);
m->addAtom(a6);
m->addAtom(a6);
m->addBond(0, 1, Bond::SINGLE);
m->addBond(1, 2, Bond::SINGLE);
m->addBond(0, 2, Bond::SINGLE);
q1 = new RWMol();
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addBond(0, 1, Bond::UNSPECIFIED);
q1->addBond(1, 2, Bond::UNSPECIFIED);
bool found = SubstructMatch(*m, *q1, matchV);
CHECK_INVARIANT(found, "");
CHECK_INVARIANT(matchV.size() == 3, "");
n = SubstructMatch(*m, *q1, matches, true);
CHECK_INVARIANT(n == 1, "");
CHECK_INVARIANT(matches[0].size() == 3, "");
// close the loop and try again (we should still match)
q1->addBond(0, 2, Bond::UNSPECIFIED);
found = SubstructMatch(*m, *q1, matchV);
CHECK_INVARIANT(found, "");
CHECK_INVARIANT(matchV.size() == 3, "");
n = SubstructMatch(*m, *q1, matches, true);
CHECK_INVARIANT(n == 1, "");
CHECK_INVARIANT(matches[0].size() == 3, "");
delete m;
delete a6;
delete q1;
std::cout << "Done\n" << std::endl;
}
void test5QueryRoot(){
std::cout << " ----------------- Test 5 QueryRoot" << std::endl;
MatchVectType matchV;
std::vector< MatchVectType > matches;
int n;
RWMol *m,*q1,*q2;
Atom *a6 = new Atom(6);
Atom *a8 = new Atom(8);
// CC(OC)C
m = new RWMol();
m->addAtom(a6);
m->addAtom(a6);
m->addAtom(a8);
m->addAtom(a6);
m->addAtom(a6);
m->addBond(0,1,Bond::SINGLE);
m->addBond(1,2,Bond::SINGLE);
m->addBond(1,4,Bond::SINGLE);
m->addBond(2,3,Bond::SINGLE);
// this will be the recursive query
q1 = new RWMol();
q1->addAtom(new QueryAtom(8),true);
q1->addAtom(new QueryAtom(6),true);
q1->addBond(0,1,Bond::UNSPECIFIED);
q1->setProp("_queryRootAtom",1);
// here's the main query
q2 = new RWMol();
QueryAtom *qA = new QueryAtom();
RecursiveStructureQuery *rsq = new RecursiveStructureQuery(q1);
qA->setQuery(rsq);
q2->addAtom(qA,true,true);
q2->addAtom(new QueryAtom(6),true,true);
q2->addBond(0,1,Bond::UNSPECIFIED);
bool found = SubstructMatch(*m,*q2,matchV);
CHECK_INVARIANT(found,"");
CHECK_INVARIANT(matchV.size()==2,"");
n = SubstructMatch(*m,*q2,matches,true);
CHECK_INVARIANT(n==2,"");
CHECK_INVARIANT(matches[0].size()==2,"");
std::cout << "Done\n" << std::endl;
}
void test2AtomMap() {
std::string rdbase = getenv("RDBASE");
std::string fname1 = rdbase + "/Code/GraphMol/MolAlign/test_data/1oir.mol";
ROMol *m1 = MolFileToMol(fname1);
std::string fname2 = rdbase + "/Code/GraphMol/MolAlign/test_data/1oir_conf.mol";
ROMol *m2 = MolFileToMol(fname2);
MatchVectType atomMap;
atomMap.push_back(std::pair<int, int>(18, 27));
atomMap.push_back(std::pair<int, int>(13, 23));
atomMap.push_back(std::pair<int, int>(21, 14));
atomMap.push_back(std::pair<int, int>(24, 7));
atomMap.push_back(std::pair<int, int>(9, 19));
atomMap.push_back(std::pair<int, int>(16, 30));
double rmsd = MolAlign::alignMol(*m2, *m1, 0, 0, &atomMap);
TEST_ASSERT(RDKit::feq(rmsd, 0.8525));
delete m1;
delete m2;
}
MatchVectType *_translateAtomMap(python::object atomMap) {
PySequenceHolder<python::object> aMapSeq(atomMap);
MatchVectType *aMap;
aMap = 0;
unsigned int i, nAtms = aMapSeq.size();
if (nAtms > 0) {
aMap = new MatchVectType;
for (i = 0; i < nAtms; ++i) {
PySequenceHolder<int> item(aMapSeq[i]);
if (item.size() != 2) {
delete aMap;
aMap = 0;
throw_value_error("Incorrect format for atomMap");
}
aMap->push_back(std::pair<int, int>(item[0], item[1]));
}
}
return aMap;
}
void test7() {
std::cout << " ----------------- Test 7 (leak check)" << std::endl;
MatchVectType matchV;
int n;
RWMol *m, *q1;
auto *a6 = new Atom(6);
bool updateLabel = true;
bool takeOwnership = true;
m = new RWMol();
m->addAtom(a6);
m->addAtom(a6);
m->addAtom(a6);
m->addBond(0, 1, Bond::SINGLE);
m->addBond(1, 2, Bond::SINGLE);
m->addBond(0, 2, Bond::SINGLE);
q1 = new RWMol();
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addBond(0, 1, Bond::UNSPECIFIED);
q1->addBond(1, 2, Bond::UNSPECIFIED);
bool found = SubstructMatch(*m, *q1, matchV);
CHECK_INVARIANT(found, "");
CHECK_INVARIANT(matchV.size() == 3, "");
std::vector<MatchVectType> matches;
for (int i = 0; i < 300000; i++) {
n = SubstructMatch(*m, *q1, matches, true, true);
CHECK_INVARIANT(n == 1, "");
CHECK_INVARIANT(matches[0].size() == 3, "");
if (!(i % 500)) std::cout << i << std::endl;
}
delete m;
delete a6;
delete q1;
std::cout << "Done\n" << std::endl;
}
bool check(MatchVectType v, MatchVectType match) {
dump("v", v);
dump("match", match);
for (size_t i = 0; i < v.size(); ++i) {
if (v[i].first != match[i].first) {
return false;
}
if (v[i].second != match[i].second) {
return false;
}
}
return true;
}
void test3Weights() {
std::string rdbase = getenv("RDBASE");
std::string fname1 = rdbase + "/Code/GraphMol/MolAlign/test_data/1oir.mol";
ROMol *m1 = MolFileToMol(fname1);
std::string fname2 = rdbase + "/Code/GraphMol/MolAlign/test_data/1oir_conf.mol";
ROMol *m2 = MolFileToMol(fname2);
MatchVectType atomMap;
atomMap.push_back(std::pair<int, int>(18, 27));
atomMap.push_back(std::pair<int, int>(13, 23));
atomMap.push_back(std::pair<int, int>(21, 14));
atomMap.push_back(std::pair<int, int>(24, 7));
atomMap.push_back(std::pair<int, int>(9, 19));
atomMap.push_back(std::pair<int, int>(16, 30));
RDNumeric::DoubleVector wts(6);
wts.setVal(0, 1.0); wts.setVal(1, 1.0);
wts.setVal(2, 1.0); wts.setVal(3, 1.0);
wts.setVal(4, 1.0); wts.setVal(5, 2.0);
double rmsd = MolAlign::alignMol(*m2, *m1, 0, 0, &atomMap, &wts);
TEST_ASSERT(RDKit::feq(rmsd, 0.9513));
delete m1;
delete m2;
}
void test8(){
std::cout << " ----------------- Test 8 (molgraph cache)" << std::endl;
MatchVectType matchV;
int n;
RWMol *m,*q1;
Atom *a6 = new Atom(6);
m = new RWMol();
m->addAtom(a6);
m->addAtom(a6);
m->addAtom(a6);
m->addBond(0,1,Bond::SINGLE);
m->addBond(1,2,Bond::SINGLE);
m->addBond(0,2,Bond::SINGLE);
q1 = new RWMol();
q1->addAtom(new QueryAtom(6),true);
q1->addAtom(new QueryAtom(6),true);
q1->addAtom(new QueryAtom(6),true);
q1->addBond(0,1,Bond::UNSPECIFIED);
q1->addBond(1,2,Bond::UNSPECIFIED);
bool found = SubstructMatch(*m,*q1,matchV);
CHECK_INVARIANT(found,"");
CHECK_INVARIANT(matchV.size()==3,"");
std::vector< MatchVectType > matches;
for(int i=0;i<30000;i++){
n = SubstructMatch(*m,*q1,matches,true,true);
CHECK_INVARIANT(n==1,"");
CHECK_INVARIANT(matches[0].size()==3,"");
if(! (i%500) ) std::cout << i << std::endl;
}
std::cout << "Done\n" << std::endl;
}
MatchVectType findFuncGroupsOnMol(const ROMol &mol,
const FragCatParams *params,
INT_VECT &fgBonds) {
PRECONDITION(params,"bad params");
fgBonds.clear();
std::pair<int, int> amat;
MatchVectType aidFgrps;
std::vector<MatchVectType> fgpMatches;
std::vector<MatchVectType>::const_iterator mati;
MatchVectType::const_iterator mi;
int aid;
//const ROMol *fgrp;
INT_VECT_CI bi;
aidFgrps.clear();
int fid = 0;
const MOL_SPTR_VECT &fgrps = params->getFuncGroups();
MOL_SPTR_VECT::const_iterator fgci;
for (fgci = fgrps.begin(); fgci != fgrps.end(); fgci++) {
const ROMol *fgrp = fgci->get();
std::string fname;
(*fgci)->getProp(common_properties::_Name, fname);
//std::cout << "Groups number: " << fname << "\n";
//(*fgci)->debugMol(std::cout);
//mol->debugMol(std::cout);
// match this functional group onto the molecule
SubstructMatch(mol, *fgrp, fgpMatches);
// loop over all the matches we get for this fgroup
for (mati = fgpMatches.begin(); mati != fgpMatches.end(); mati++) {
//FIX: we will assume that the first atom in fgrp is always the connection
// atom
amat = mati->front();
aid = amat.second; //FIX: is this correct - the second entry in the pair is the atom ID from mol
// grab the list of atom Ids from mol that match the functional group
INT_VECT bondIds, maids;
for (mi = mati->begin(); mi != mati->end(); mi++) {
maids.push_back(mi->second);
}
// create a list of bond IDs from these atom ID
// these are the bond in mol that are part of portion that matches the
// functional group
bondIds = Subgraphs::bondListFromAtomList(mol, maids);
// now check if all these bonds have been covered as part of larger
// functional group that was dealt with earlier
// FIX: obviously we assume here that the function groups in params
// come in decreasing order of size.
bool allDone = true;
for (bi = bondIds.begin(); bi != bondIds.end(); bi++) {
if (std::find(fgBonds.begin(), fgBonds.end(), (*bi)) == fgBonds.end()) {
allDone = false;
fgBonds.push_back(*bi);
}
}
if (!allDone) {
// this functional group mapping onto mol is not part of a larger func
// group mapping so record it
aidFgrps.push_back(std::pair<int, int>(aid, fid));
}
}
fid++;
}
return aidFgrps;
}
void testGitHubIssue688() {
BOOST_LOG(rdErrorLog) << "-------------------------------------" << std::endl;
BOOST_LOG(rdErrorLog) << " Test GitHub issue 688: partially specified "
"chiral substructure queries don't work properly"
<< std::endl;
{
std::string smi = "C1CC[C@](Cl)(N)O1";
ROMol *mol = SmilesToMol(smi);
TEST_ASSERT(mol);
// mol->debugMol(std::cerr);
std::string sma = "C1CC[C@](N)O1";
ROMol *qmol = SmartsToMol(sma);
TEST_ASSERT(qmol);
// qmol->updatePropertyCache();
// qmol->debugMol(std::cerr);
MatchVectType match;
TEST_ASSERT(SubstructMatch(*mol, *qmol, match, true, false));
TEST_ASSERT(match.size() == qmol->getNumAtoms());
TEST_ASSERT(SubstructMatch(*mol, *qmol, match, true, true));
TEST_ASSERT(match.size() == qmol->getNumAtoms());
delete mol;
delete qmol;
}
{
std::string smi = "C1CC[C@](Cl)(N)O1";
ROMol *mol = SmilesToMol(smi);
TEST_ASSERT(mol);
// mol->debugMol(std::cerr);
std::string sma = "C1CC[C@@](N)O1";
ROMol *qmol = SmartsToMol(sma);
TEST_ASSERT(qmol);
// qmol->updatePropertyCache();
// qmol->debugMol(std::cerr);
MatchVectType match;
TEST_ASSERT(SubstructMatch(*mol, *qmol, match, true, false));
TEST_ASSERT(match.size() == qmol->getNumAtoms());
TEST_ASSERT(!SubstructMatch(*mol, *qmol, match, true, true));
delete mol;
delete qmol;
}
{
std::string smi = "N[C@]1(Cl)CCCO1";
ROMol *mol = SmilesToMol(smi);
TEST_ASSERT(mol);
// mol->debugMol(std::cerr);
std::string sma = "N[C@]1CCCO1";
ROMol *qmol = SmartsToMol(sma);
TEST_ASSERT(qmol);
// qmol->updatePropertyCache();
// qmol->debugMol(std::cerr);
// std::cerr << MolToSmiles(*qmol, true) << std::endl;
MatchVectType match;
TEST_ASSERT(SubstructMatch(*mol, *qmol, match, true, false));
TEST_ASSERT(match.size() == qmol->getNumAtoms());
TEST_ASSERT(SubstructMatch(*mol, *qmol, match, true, true));
TEST_ASSERT(match.size() == qmol->getNumAtoms());
delete mol;
delete qmol;
}
{
std::string smi = "N[C@]1(Cl)CCCO1";
ROMol *mol = SmilesToMol(smi);
TEST_ASSERT(mol);
// mol->debugMol(std::cerr);
std::string sma = "N[C@@]1CCCO1";
ROMol *qmol = SmartsToMol(sma);
TEST_ASSERT(qmol);
// qmol->updatePropertyCache();
// qmol->debugMol(std::cerr);
// std::cerr << MolToSmiles(*qmol, true) << std::endl;
MatchVectType match;
TEST_ASSERT(SubstructMatch(*mol, *qmol, match, true, false));
TEST_ASSERT(match.size() == qmol->getNumAtoms());
TEST_ASSERT(!SubstructMatch(*mol, *qmol, match, true, true));
delete mol;
delete qmol;
}
BOOST_LOG(rdErrorLog) << " done" << std::endl;
}
void test5() {
BOOST_LOG(rdErrorLog) << "-------------------------------------" << std::endl;
BOOST_LOG(rdErrorLog) << "FeatureFactory testing." << std::endl;
ROMol *testMol;
MatchVectType mv;
std::string inText;
FeatSPtrList featSPtrs;
boost::shared_ptr<MolChemicalFeature> featSPtr;
MolChemicalFeatureFactory *factory;
MolChemicalFeatureDef::CollectionType::value_type featDef;
inText =
"DefineFeature HDonor1 [N,O;!H0]\n"
" Family HBondDonor\n"
" Weights 1.0\n"
"EndFeature\n"
"DefineFeature HAcceptor1 [N,O]\n"
" Family HBondAcceptor\n"
" Weights 1.0\n"
"EndFeature\n";
factory = buildFeatureFactory(inText);
TEST_ASSERT(factory);
TEST_ASSERT(factory->getNumFeatureDefs() == 2);
testMol = SmilesToMol("COCN");
TEST_ASSERT(testMol);
featDef = *factory->beginFeatureDefs();
TEST_ASSERT(SubstructMatch(*testMol, *featDef->getPattern(), mv));
BOOST_LOG(rdErrorLog) << "1" << std::endl;
TEST_ASSERT(mv.size() == 1);
TEST_ASSERT(mv[0].first == 0);
TEST_ASSERT(mv[0].second == 3);
featDef = *(++factory->beginFeatureDefs());
TEST_ASSERT(SubstructMatch(*testMol, *featDef->getPattern(), mv));
BOOST_LOG(rdErrorLog) << "2" << std::endl;
TEST_ASSERT(mv.size() == 1);
TEST_ASSERT(mv[0].first == 0);
TEST_ASSERT(mv[0].second == 1 || mv[0].second == 3);
// Test using the factory to find features:
featSPtrs = factory->getFeaturesForMol(*testMol);
BOOST_LOG(rdErrorLog) << "3" << std::endl;
TEST_ASSERT(featSPtrs.size() == 3);
featSPtr = *featSPtrs.begin();
TEST_ASSERT(featSPtr->getFamily() == "HBondDonor");
TEST_ASSERT(featSPtr->getType() == "HDonor1");
featSPtr = *(++featSPtrs.begin());
TEST_ASSERT(featSPtr->getFamily() == "HBondAcceptor");
TEST_ASSERT(featSPtr->getType() == "HAcceptor1");
featSPtr = *(++featSPtrs.begin());
TEST_ASSERT(featSPtr->getFamily() == "HBondAcceptor");
TEST_ASSERT(featSPtr->getType() == "HAcceptor1");
// Test limiting stuff with includeOnly
featSPtrs = factory->getFeaturesForMol(*testMol, "HBondAcceptor");
BOOST_LOG(rdErrorLog) << "4" << std::endl;
TEST_ASSERT(featSPtrs.size() == 2);
featSPtr = *featSPtrs.begin();
TEST_ASSERT(featSPtr->getFamily() == "HBondAcceptor");
TEST_ASSERT(featSPtr->getType() == "HAcceptor1");
featSPtr = *(++featSPtrs.begin());
TEST_ASSERT(featSPtr->getFamily() == "HBondAcceptor");
TEST_ASSERT(featSPtr->getType() == "HAcceptor1");
featSPtrs = factory->getFeaturesForMol(*testMol, "HBondDonor");
BOOST_LOG(rdErrorLog) << "5" << std::endl;
TEST_ASSERT(featSPtrs.size() == 1);
featSPtr = *featSPtrs.begin();
TEST_ASSERT(featSPtr->getFamily() == "HBondDonor");
TEST_ASSERT(featSPtr->getType() == "HDonor1");
featSPtrs = factory->getFeaturesForMol(*testMol, "NotPresent");
BOOST_LOG(rdErrorLog) << "6" << std::endl;
TEST_ASSERT(featSPtrs.size() == 0);
delete testMol;
delete factory;
BOOST_LOG(rdErrorLog) << " done" << std::endl;
}
void test3() {
std::cout << " ----------------- Test 3" << std::endl;
MatchVectType matchV;
std::vector<MatchVectType> matches;
unsigned int n;
RWMol *m, *q1;
m = new RWMol();
bool updateLabel = true;
bool takeOwnership = true;
m->addAtom(new Atom(6), updateLabel, takeOwnership);
m->addAtom(new Atom(6), updateLabel, takeOwnership);
m->addAtom(new Atom(8), updateLabel, takeOwnership);
m->addBond(0, 1, Bond::SINGLE);
m->addBond(1, 2, Bond::SINGLE);
q1 = new RWMol();
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addAtom(new QueryAtom(8), updateLabel, takeOwnership);
q1->addBond(0, 1, Bond::UNSPECIFIED);
n = SubstructMatch(*m, *q1, matches, false);
CHECK_INVARIANT(n == 1, "");
CHECK_INVARIANT(matches.size() == n, "");
CHECK_INVARIANT(matches[0].size() == 2, "");
n = SubstructMatch(*m, *q1, matches, true);
CHECK_INVARIANT(n == 1, "");
CHECK_INVARIANT(matches.size() == n, "");
CHECK_INVARIANT(matches[0].size() == 2, "");
CHECK_INVARIANT(SubstructMatch(*m, *q1, matchV), "");
CHECK_INVARIANT(matchV.size() == 2, "");
delete m;
m = new RWMol();
m->addAtom(new Atom(6), updateLabel, takeOwnership);
m->addAtom(new Atom(6), updateLabel, takeOwnership);
m->addAtom(new Atom(8), updateLabel, takeOwnership);
m->addBond(0, 1, Bond::SINGLE);
m->addBond(1, 2, Bond::DOUBLE);
matches.clear();
n = SubstructMatch(*m, *q1, matches, false);
CHECK_INVARIANT(n == 1, "");
CHECK_INVARIANT(matches.size() == n, "");
CHECK_INVARIANT(matches[0].size() == 2, "");
n = SubstructMatch(*m, *q1, matches, true);
CHECK_INVARIANT(n == 1, "");
CHECK_INVARIANT(matches.size() == n, "");
CHECK_INVARIANT(matches[0].size() == 2, "");
CHECK_INVARIANT(SubstructMatch(*m, *q1, matchV), "");
CHECK_INVARIANT(matchV.size() == 2, "");
delete q1;
q1 = new RWMol();
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addAtom(new QueryAtom(6), updateLabel, takeOwnership);
q1->addBond(0, 1, Bond::UNSPECIFIED);
n = SubstructMatch(*m, *q1, matches, false);
TEST_ASSERT(n == 2);
TEST_ASSERT(matches.size() == n);
TEST_ASSERT(matches[0].size() == 2);
TEST_ASSERT(matches[1].size() == 2);
TEST_ASSERT(matches[0][0].second != matches[1][0].second);
TEST_ASSERT(matches[0][1].second != matches[1][1].second);
n = SubstructMatch(*m, *q1, matches, true);
TEST_ASSERT(n == 1);
TEST_ASSERT(matches.size() == n);
delete m;
delete q1;
std::cout << "Done\n" << std::endl;
}
void test1(){
std::cout << " ----------------- Test 1" << std::endl;
MatchVectType matchV;
std::vector< MatchVectType > matches;
unsigned int n;
RWMol *m,*q1;
m = new RWMol();
m->addAtom(new Atom(8));
m->addAtom(new Atom(6));
m->addAtom(new Atom(6));
m->addBond(0,1,Bond::SINGLE);
m->addBond(1,2,Bond::SINGLE);
q1 = new RWMol();
q1->addAtom(new QueryAtom(6));
q1->addAtom(new QueryAtom(6));
q1->addBond(0,1,Bond::SINGLE);
n = SubstructMatch(*m,*q1,matches,false);
CHECK_INVARIANT(n==2,"");
CHECK_INVARIANT(matches.size()==n,"");
CHECK_INVARIANT(matches[0].size()==2,"");
TEST_ASSERT(matches[0][0].first==0);
TEST_ASSERT(matches[0][0].second==1||matches[0][0].second==2);
TEST_ASSERT(matches[0][1].first==1);
TEST_ASSERT(matches[0][1].second!=matches[0][0].second);
TEST_ASSERT(matches[1][1].second==1||matches[0][1].second==2);
TEST_ASSERT(matches[1][0].first==0);
TEST_ASSERT(matches[1][0].second==1||matches[1][0].second==2);
TEST_ASSERT(matches[1][0].second!=matches[0][0].second);
TEST_ASSERT(matches[1][0].second==matches[0][1].second);
TEST_ASSERT(matches[1][1].first==1);
TEST_ASSERT(matches[1][1].second!=matches[1][0].second);
TEST_ASSERT(matches[1][1].second==matches[0][0].second);
n = SubstructMatch(*m,*q1,matches,true);
CHECK_INVARIANT(n==1,"");
CHECK_INVARIANT(matches.size()==n,"");
CHECK_INVARIANT(matches[0].size()==2,"");
TEST_ASSERT(matches[0][0].first==0);
TEST_ASSERT(matches[0][0].second==1||matches[0][0].second==2);
TEST_ASSERT(matches[0][1].first==1);
TEST_ASSERT(matches[0][1].second!=matches[0][0].second);
TEST_ASSERT(matches[0][1].second==1||matches[0][1].second==2);
CHECK_INVARIANT(SubstructMatch(*m,*q1,matchV),"");
CHECK_INVARIANT(matchV.size()==2,"");
// make sure we reset the match vectors.
// build a query we won't match:
q1->addAtom(new QueryAtom(6));
q1->addBond(1,2,Bond::SINGLE);
q1->addAtom(new QueryAtom(6));
q1->addBond(2,3,Bond::SINGLE);
TEST_ASSERT(!SubstructMatch(*m,*q1,matchV));
TEST_ASSERT(matchV.size()==0);
n = SubstructMatch(*m,*q1,matches,false);
TEST_ASSERT(n==0);
TEST_ASSERT(matches.size()==0);
std::cout << "Done\n" << std::endl;
}
void addTorsions(const ROMol &mol, MMFFMolProperties *mmffMolProperties,
ForceFields::ForceField *field,
const std::string &torsionBondSmarts) {
PRECONDITION(field, "bad ForceField");
PRECONDITION(mmffMolProperties, "bad MMFFMolProperties");
PRECONDITION(mmffMolProperties->isValid(),
"missing atom types - invalid force-field");
std::ostream &oStream = mmffMolProperties->getMMFFOStream();
ROMol::ADJ_ITER nbr1Idx;
ROMol::ADJ_ITER end1Nbrs;
ROMol::ADJ_ITER nbr2Idx;
ROMol::ADJ_ITER end2Nbrs;
double totalTorsionEnergy = 0.0;
RDGeom::PointPtrVect points;
if (mmffMolProperties->getMMFFVerbosity()) {
if (mmffMolProperties->getMMFFVerbosity() == MMFF_VERBOSITY_HIGH) {
oStream << "\n"
"T O R S I O N A L\n\n"
"--------------ATOMS--------------- ---ATOM TYPES--- "
"FF TORSION -----FORCE Params-----\n"
" I J K L I J K L "
"CLASS ANGLE ENERGY V1 V2 V3\n"
"-------------------------------------------------------------"
"-----------------------------------------------------"
<< std::endl;
}
points = field->positions();
}
std::vector<MatchVectType> matchVect;
const ROMol *defaultQuery = DefaultTorsionBondSmarts::query();
const ROMol *query = (torsionBondSmarts == DefaultTorsionBondSmarts::string())
? defaultQuery : SmartsToMol(torsionBondSmarts);
TEST_ASSERT(query);
unsigned int nHits = SubstructMatch(mol, *query, matchVect);
if (query != defaultQuery) delete query;
for (unsigned int i = 0; i < nHits; ++i) {
MatchVectType match = matchVect[i];
TEST_ASSERT(match.size() == 2);
int idx2 = match[0].second;
int idx3 = match[1].second;
const Bond *bond = mol.getBondBetweenAtoms(idx2, idx3);
TEST_ASSERT(bond);
const Atom *jAtom = mol.getAtomWithIdx(idx2);
const Atom *kAtom = mol.getAtomWithIdx(idx3);
if (((jAtom->getHybridization() == Atom::SP2) ||
(jAtom->getHybridization() == Atom::SP3)) &&
((kAtom->getHybridization() == Atom::SP2) ||
(kAtom->getHybridization() == Atom::SP3))) {
ROMol::OEDGE_ITER beg1, end1;
boost::tie(beg1, end1) = mol.getAtomBonds(jAtom);
while (beg1 != end1) {
const Bond *tBond1 = mol[*beg1].get();
if (tBond1 != bond) {
int idx1 = tBond1->getOtherAtomIdx(idx2);
ROMol::OEDGE_ITER beg2, end2;
boost::tie(beg2, end2) = mol.getAtomBonds(kAtom);
while (beg2 != end2) {
const Bond *tBond2 = mol[*beg2].get();
if ((tBond2 != bond) && (tBond2 != tBond1)) {
int idx4 = tBond2->getOtherAtomIdx(idx3);
// make sure this isn't a three-membered ring:
if (idx4 != idx1) {
// we now have a torsion involving atoms (bonds):
// bIdx - (tBond1) - idx1 - (bond) - idx2 - (tBond2) - eIdx
unsigned int torType;
MMFFTor mmffTorParams;
if (mmffMolProperties->getMMFFTorsionParams(
mol, idx1, idx2, idx3, idx4, torType, mmffTorParams)) {
TorsionAngleContrib *contrib = new TorsionAngleContrib(
field, idx1, idx2, idx3, idx4, &mmffTorParams);
field->contribs().push_back(ForceFields::ContribPtr(contrib));
if (mmffMolProperties->getMMFFVerbosity()) {
const Atom *iAtom = mol.getAtomWithIdx(idx1);
const Atom *lAtom = mol.getAtomWithIdx(idx4);
unsigned int iAtomType =
mmffMolProperties->getMMFFAtomType(idx1);
unsigned int jAtomType =
mmffMolProperties->getMMFFAtomType(idx2);
unsigned int kAtomType =
mmffMolProperties->getMMFFAtomType(idx3);
unsigned int lAtomType =
mmffMolProperties->getMMFFAtomType(idx4);
const RDGeom::Point3D p1((*(points[idx1]))[0],
(*(points[idx1]))[1],
(*(points[idx1]))[2]);
const RDGeom::Point3D p2((*(points[idx2]))[0],
(*(points[idx2]))[1],
(*(points[idx2]))[2]);
const RDGeom::Point3D p3((*(points[idx3]))[0],
(*(points[idx3]))[1],
(*(points[idx3]))[2]);
const RDGeom::Point3D p4((*(points[idx4]))[0],
(*(points[idx4]))[1],
(*(points[idx4]))[2]);
const double cosPhi =
MMFF::Utils::calcTorsionCosPhi(p1, p2, p3, p4);
const double torsionEnergy = MMFF::Utils::calcTorsionEnergy(
mmffTorParams.V1, mmffTorParams.V2, mmffTorParams.V3,
cosPhi);
if (mmffMolProperties->getMMFFVerbosity() ==
MMFF_VERBOSITY_HIGH) {
oStream
<< std::left << std::setw(2) << iAtom->getSymbol()
<< " #" << std::setw(5) << idx1 + 1 << std::setw(2)
<< jAtom->getSymbol() << " #" << std::setw(5)
<< idx2 + 1 << std::setw(2) << kAtom->getSymbol()
<< " #" << std::setw(5) << idx3 + 1 << std::setw(2)
<< lAtom->getSymbol() << " #" << std::setw(5)
<< idx4 + 1 << std::right << std::setw(5) << iAtomType
<< std::setw(5) << jAtomType << std::setw(5)
<< kAtomType << std::setw(5) << lAtomType
<< std::setw(6) << torType << " " << std::fixed
<< std::setprecision(3) << std::setw(10)
<< RAD2DEG * acos(cosPhi) << std::setw(10)
<< torsionEnergy << std::setw(10) << mmffTorParams.V1
<< std::setw(10) << mmffTorParams.V2 << std::setw(10)
<< mmffTorParams.V3 << std::endl;
}
totalTorsionEnergy += torsionEnergy;
}
}
}
}
beg2++;
}
}
beg1++;
}
}
}
if (mmffMolProperties->getMMFFVerbosity()) {
if (mmffMolProperties->getMMFFVerbosity() == MMFF_VERBOSITY_HIGH) {
oStream << std::endl;
}
oStream << "TOTAL TORSIONAL ENERGY =" << std::right << std::setw(16)
<< std::fixed << std::setprecision(4) << totalTorsionEnergy
<< std::endl;
}
}
