private Map<IAtom, List<String>> labelAtomsBySymbol(IAtomContainer atomCont) {
Map<IAtom, List<String>> label_list = new HashMap<>();
for (int i = 0; i < atomCont.getAtomCount(); i++) {
List<String> label = new ArrayList<>(7);
for (int a = 0; a < 7; a++) {
label.add(a, "Z9");
}
IAtom refAtom = atomCont.getAtom(i);
/*
* Important Step: Discriminate between source atom types
*/
String referenceAtom;
if (refAtom instanceof IQueryAtom) {
referenceAtom =
((IQueryAtom) refAtom).getSymbol() == null ? "*" : ((IQueryAtom) refAtom).getSymbol();
// System.out.println("referenceAtom " + referenceAtom);
} else if (!(refAtom instanceof IQueryAtom) && this.matchAtomType) {
referenceAtom =
refAtom.getAtomTypeName() == null ? refAtom.getSymbol() : refAtom.getAtomTypeName();
} else {
referenceAtom = refAtom.getSymbol();
}
label.set(0, referenceAtom);
List<IAtom> connAtoms = atomCont.getConnectedAtomsList(refAtom);
int counter = 1;
for (IAtom negAtom : connAtoms) {
String neighbouringAtom;
if (refAtom instanceof IQueryAtom) {
neighbouringAtom =
((IQueryAtom) negAtom).getSymbol() == null ? "*" : ((IQueryAtom) negAtom).getSymbol();
// System.out.println("neighbouringAtom " + neighbouringAtom);
} else if (!(negAtom instanceof IQueryAtom) && this.matchAtomType) {
neighbouringAtom =
negAtom.getAtomTypeName() == null ? negAtom.getSymbol() : negAtom.getAtomTypeName();
} else {
neighbouringAtom = negAtom.getSymbol();
}
label.set(counter, neighbouringAtom);
counter += 1;
}
// System.out.println("label " + label);
bubbleSort(label);
label_list.put(refAtom, label);
}
return label_list;
}
@Test
public void testSybylAtomTypePerceptionBenzene()
throws CDKException, FileNotFoundException, IOException, BioclipseException, CoreException,
InvocationTargetException {
IAtomContainer ac = MoleculeFactory.makeBenzene();
ICDKMolecule mol = new CDKMolecule(ac);
debug.perceiveSybylAtomTypes(mol);
System.out.println("** BENZENE **");
System.out.println(AtomContainerDiff.diff(ac, mol.getAtomContainer()));
for (int i = 0; i < mol.getAtomContainer().getAtomCount(); i++) {
IAtom a = mol.getAtomContainer().getAtom(i);
System.out.println("Atom: " + a.getSymbol() + i + ", type=" + a.getAtomTypeName());
}
assertEquals("C.ar", mol.getAtomContainer().getAtom(0).getAtomTypeName());
assertEquals("C.ar", mol.getAtomContainer().getAtom(1).getAtomTypeName());
assertEquals("C.ar", mol.getAtomContainer().getAtom(2).getAtomTypeName());
assertEquals("C.ar", mol.getAtomContainer().getAtom(3).getAtomTypeName());
assertEquals("C.ar", mol.getAtomContainer().getAtom(4).getAtomTypeName());
assertEquals("C.ar", mol.getAtomContainer().getAtom(5).getAtomTypeName());
}
/** Converts an Atom to a MSML Atom element */
protected AtomType convertAtom(IAtom atom, String parentID) {
AtomType atomElement = new AtomType();
String id = "";
if (atom.getID() != null) {
id = atom.getID();
} else {
id = Integer.toString(atom.hashCode());
}
String atomID = parentID + PREFIX_ATOM + id;
atomElement.setId(atomID);
atomElement.setCustomId("" + id);
// atom name
atomElement.setTitle(atom.getAtomTypeName());
// element name
atomElement.setElementType(atom.getSymbol());
double x, y, z;
// set choords
if (atom.getPoint3d() != null) {
x = atom.getPoint3d().x;
y = atom.getPoint3d().y;
z = atom.getPoint3d().z;
} else { // what if mol in 2d? -> has to use getPoint2d
x = atom.getPoint2d().x;
y = atom.getPoint2d().y;
z = 0.0;
}
if (atom.getFormalCharge() != null) {
atomElement.setFormalCharge(BigInteger.valueOf(atom.getFormalCharge()));
}
atomElement.setX3(new Float(x));
atomElement.setY3(new Float(y));
atomElement.setZ3(new Float(z));
return atomElement;
}
/**
* This method calculates the ionization potential of an atom.
*
* @param atom The IAtom to ionize.
* @param container Parameter is the IAtomContainer.
* @return The ionization potential. Not possible the ionization.
*/
@Override
public DescriptorValue calculate(IAtom atom, IAtomContainer container) {
double value = 0;
// FIXME: for now I'll cache a few modified atomic properties, and restore them at the end of
// this method
String originalAtomtypeName = atom.getAtomTypeName();
Integer originalNeighborCount = atom.getFormalNeighbourCount();
Integer originalValency = atom.getValency();
IAtomType.Hybridization originalHybrid = atom.getHybridization();
Double originalBondOrderSum = atom.getBondOrderSum();
Order originalMaxBondOrder = atom.getMaxBondOrder();
if (!isCachedAtomContainer(container)) {
try {
AtomContainerManipulator.percieveAtomTypesAndConfigureAtoms(container);
LonePairElectronChecker lpcheck = new LonePairElectronChecker();
lpcheck.saturate(container);
} catch (CDKException e) {
return new DescriptorValue(
getSpecification(),
getParameterNames(),
getParameters(),
new DoubleResult(Double.NaN),
getDescriptorNames(),
e);
}
}
try {
value = IonizationPotentialTool.predictIP(container, atom);
} catch (CDKException e) {
return new DescriptorValue(
getSpecification(),
getParameterNames(),
getParameters(),
new DoubleResult(Double.NaN),
getDescriptorNames(),
e);
}
// restore original props
atom.setAtomTypeName(originalAtomtypeName);
atom.setFormalNeighbourCount(originalNeighborCount);
atom.setValency(originalValency);
atom.setHybridization(originalHybrid);
atom.setMaxBondOrder(originalMaxBondOrder);
atom.setBondOrderSum(originalBondOrderSum);
return new DescriptorValue(
getSpecification(),
getParameterNames(),
getParameters(),
new DoubleResult(value),
getDescriptorNames());
}
/** Tests if the electron count matches the Hückel 4n+2 rule. */
private static boolean isHueckelValid(IAtomContainer singleRing) throws CDKException {
int electronCount = 0;
for (IAtom ringAtom : singleRing.atoms()) {
if (ringAtom.getHybridization() != CDKConstants.UNSET
&& (ringAtom.getHybridization() == Hybridization.SP2)
|| ringAtom.getHybridization() == Hybridization.PLANAR3) {
// for example, a carbon
// note: the double bond is in the ring, that has been tested earlier
// FIXME: this does assume bond orders to be resolved too, when detecting
// sprouting double bonds
if ("N.planar3".equals(ringAtom.getAtomTypeName())) {
electronCount += 2;
} else if ("N.minus.planar3".equals(ringAtom.getAtomTypeName())) {
electronCount += 2;
} else if ("N.amide".equals(ringAtom.getAtomTypeName())) {
electronCount += 2;
} else if ("S.2".equals(ringAtom.getAtomTypeName())) {
electronCount += 2;
} else if ("S.planar3".equals(ringAtom.getAtomTypeName())) {
electronCount += 2;
} else if ("C.minus.planar".equals(ringAtom.getAtomTypeName())) {
electronCount += 2;
} else if ("O.planar3".equals(ringAtom.getAtomTypeName())) {
electronCount += 2;
} else if ("N.sp2.3".equals(ringAtom.getAtomTypeName())) {
electronCount += 1;
} else {
if (factory == null) {
factory =
AtomTypeFactory.getInstance(
"org/openscience/cdk/dict/data/cdk-atom-types.owl", ringAtom.getBuilder());
}
IAtomType type = factory.getAtomType(ringAtom.getAtomTypeName());
Object property = type.getProperty(CDKConstants.PI_BOND_COUNT);
if (property != null && property instanceof Integer) {
electronCount += ((Integer) property).intValue();
}
}
} else if (ringAtom.getHybridization() != null
&& ringAtom.getHybridization() == Hybridization.SP3
&& getLonePairCount(ringAtom) > 0) {
// for example, a nitrogen or oxygen
electronCount += 2;
}
}
return (electronCount % 4 == 2) && (electronCount > 2);
}
@Test
public void testSybylAtomTypePerceptionFromSMILES()
throws FileNotFoundException, IOException, BioclipseException, CoreException,
InvocationTargetException {
ICDKMolecule mol = cdk.fromSMILES("C1CCCCC1CCOC");
debug.perceiveSybylAtomTypes(mol);
for (int i = 0; i < mol.getAtomContainer().getAtomCount(); i++) {
IAtom a = mol.getAtomContainer().getAtom(i);
System.out.println("Atom: " + a.getSymbol() + i + ", type=" + a.getAtomTypeName());
}
}
@Test
public void testSybylAtomTypePerception()
throws FileNotFoundException, IOException, BioclipseException, CoreException,
InvocationTargetException {
String path = getClass().getResource("/testFiles/atp.mol").getPath();
ICDKMolecule mol = cdk.loadMolecule(new MockIFile(path));
System.out.println("mol: " + mol.toString());
debug.perceiveSybylAtomTypes(mol);
for (int i = 0; i < mol.getAtomContainer().getAtomCount(); i++) {
IAtom a = mol.getAtomContainer().getAtom(i);
System.out.println("Atom: " + a.getSymbol() + i + ", type=" + a.getAtomTypeName());
}
}
/**
* Determines if the isolatedRingSystem has attached double bonds, which are not part of the ring
* system itself, and not part of any other ring system. Exceptions: a N.sp2.3 nitrogen with a
* double ring to an oxygen outwards.
*/
private static boolean isRingSystemSproutedWithNonRingDoubleBonds(
IAtomContainer fullContainer, IAtomContainer isolatedRingSystem) {
Iterator<IAtom> atoms = isolatedRingSystem.atoms().iterator();
while (atoms.hasNext()) {
IAtom atom = atoms.next();
Iterator<IBond> neighborBonds = fullContainer.getConnectedBondsList(atom).iterator();
while (neighborBonds.hasNext()) {
IBond neighborBond = neighborBonds.next();
if (!neighborBond.getFlag(CDKConstants.ISINRING)
&& neighborBond.getOrder() == CDKConstants.BONDORDER_DOUBLE
|| neighborBond.getOrder() == CDKConstants.BONDORDER_TRIPLE) {
if (!("N.sp2.3".equals(atom.getAtomTypeName())
&& "O.sp2".equals(neighborBond.getConnectedAtom(atom).getAtomTypeName())))
return true;
}
}
}
return false;
}
public String perceiveSybylAtomTypes(IMolecule mol)
throws InvocationTargetException {
ICDKMolecule cdkmol;
try {
cdkmol = cdk.asCDKMolecule(mol);
}
catch (BioclipseException e) {
System.out.println("Error converting cdk10 to cdk");
e.printStackTrace();
throw new InvocationTargetException(e);
}
IAtomContainer ac = cdkmol.getAtomContainer();
CDKAtomTypeMatcher cdkMatcher
= CDKAtomTypeMatcher.getInstance(ac.getBuilder());
AtomTypeMapper mapper
= AtomTypeMapper.getInstance(
"org/openscience/cdk/dict/data/cdk-sybyl-mappings.owl" );
IAtomType[] sybylTypes = new IAtomType[ac.getAtomCount()];
int atomCounter = 0;
int a=0;
for (IAtom atom : ac.atoms()) {
IAtomType type;
try {
type = cdkMatcher.findMatchingAtomType(ac, atom);
}
catch (CDKException e) {
type = null;
}
if (type==null) {
// logger.debug("AT null for atom: " + atom);
type = atom.getBuilder().newAtomType(atom.getSymbol());
type.setAtomTypeName("X");
}
AtomTypeManipulator.configure(atom, type);
a++;
}
try {
CDKHueckelAromaticityDetector.detectAromaticity(ac);
// System.out.println("Arom: "
// + CDKHueckelAromaticityDetector.detectAromaticity(ac) );
}
catch (CDKException e) {
logger.debug("Failed to perceive aromaticity: " + e.getMessage());
}
for (IAtom atom : ac.atoms()) {
String mappedType = mapper.mapAtomType(atom.getAtomTypeName());
if ("C.2".equals(mappedType)
&& atom.getFlag(CDKConstants.ISAROMATIC)) {
mappedType = "C.ar";
}
else if ("N.pl3".equals(mappedType)
&& atom.getFlag(CDKConstants.ISAROMATIC)) {
mappedType = "N.ar";
}
try {
sybylTypes[atomCounter] = factory.getAtomType(mappedType);
}
catch (NoSuchAtomTypeException e) {
// yes, setting null's here is important
sybylTypes[atomCounter] = null;
}
atomCounter++;
}
StringBuffer result = new StringBuffer();
// now that full perception is finished, we can set atom type names:
for (int i = 0; i < sybylTypes.length; i++) {
if (sybylTypes[i] != null) {
ac.getAtom(i).setAtomTypeName(sybylTypes[i].getAtomTypeName());
}
else {
ac.getAtom(i).setAtomTypeName("X");
}
result.append(i).append(':').append(ac.getAtom(i).getAtomTypeName())
/*.append("\n")*/;
}
return result.toString();
}
/**
* @return
* @throws Exception
*/
public int process() throws Exception {
if (file == null) throw new Exception("File not assigned! Use -f command line option.");
if (!file.exists()) throw new FileNotFoundException(file.getAbsolutePath());
int records_read = 0;
int records_processed = 0;
int records_error = 0;
InputStream in = new FileInputStream(file);
/**
* cdk-io module http://ambit.uni-plovdiv.bg:8083/nexus/index.html#nexus-
* search;classname~IteratingMDLReader
*/
IteratingSDFReader reader = null;
SDFWriter writer = new SDFWriter(new OutputStreamWriter(System.out));
try {
reader = new IteratingSDFReader(in, DefaultChemObjectBuilder.getInstance());
LOGGER.log(Level.INFO, String.format("Reading %s", file.getAbsoluteFile()));
while (reader.hasNext()) {
/** Note recent versions allow IAtomContainer molecule = reader.next(); */
Object object = reader.next();
IAtomContainer molecule = null;
if (object instanceof IAtomContainer) molecule = (IAtomContainer) object;
else break;
records_read++;
try {
/** cdk-standard module */
AtomContainerManipulator.percieveAtomTypesAndConfigureAtoms(molecule);
// CDKHueckelAromaticityDetector.detectAromaticity(molecule);
for (IAtom atom : molecule.atoms())
if (atom.getImplicitHydrogenCount() == null) {
LOGGER.fine(
atom.getSymbol()
+ "\t"
+ atom.getAtomTypeName()
+ "\t"
+ atom.getImplicitHydrogenCount());
atom.setImplicitHydrogenCount(0);
}
molecule = AtomContainerManipulator.copyAndSuppressedHydrogens(molecule);
/** Generate SMILES and assign as properties */
assignSMILES(molecule);
molecule.setProperty("REACTION", "REACTANT");
molecule.setProperty("SMIRKS", "");
/** Apply reactions */
writer.write(molecule);
for (int r = 0; r < smirks.length; r++) {
if (reactions[r] == null) {
reactions[r] = smrkMan.parse(smirks[r][1]);
}
IAtomContainer reactant = molecule.clone();
if (smrkMan.applyTransformation(reactant, reactions[r])) {
AtomContainerManipulator.percieveAtomTypesAndConfigureAtoms(reactant);
reactant.setProperty("REACTION", "PRODUCT OF " + smirks[r][0]);
reactant.setProperty("SMIRKS", smirks[r][1]);
try {
assignSMILES(reactant);
} catch (Exception x) {
LOGGER.log(Level.WARNING, x.getMessage());
}
writer.write(reactant);
}
}
records_processed++;
;
} catch (Exception x) {
System.err.println("*");
records_error++;
LOGGER.log(
Level.SEVERE,
String.format("[Record %d] Error %s\n", records_read, file.getAbsoluteFile()),
x);
}
}
} catch (Exception x) {
LOGGER.log(
Level.SEVERE,
String.format("[Record %d] Error %s\n", records_read, file.getAbsoluteFile()),
x);
} finally {
try {
reader.close();
} catch (Exception x) {
}
try {
writer.close();
} catch (Exception x) {
}
}
LOGGER.log(
Level.INFO,
String.format(
"[Records read/processed/error %d/%d/%d] %s",
records_read, records_processed, records_error, file.getAbsoluteFile()));
return records_read;
}