public static IAtomContainer GetLargestMoleculeFragment(
IAtomContainer mol,
out int acc)
{
IAtomContainer mol2, molMain = null;
acc = 0;
if (mol == null)
{
return(null);
}
if (ConnectivityChecker.IsConnected(mol))
{
acc = 1;
return(mol);
}
IReadOnlyList <IAtomContainer> acs = ConnectivityChecker.PartitionIntoMolecules(mol);
int largestAc = -1;
acc = acs.Count;
for (int aci = 0; aci < acc; aci++)
{
mol2 = acs[aci];
int ac2 = mol2.Atoms.Count;
if (ac2 > largestAc)
{
largestAc = mol2.Atoms.Count;
molMain = mol2;
}
}
return(molMain);
}
/// <summary>
/// </summary>
/// <param name="mol"></param>
/// <param name="mcss"></param>
/// <param name="shouldMatchBonds"></param>
/// <returns>IMolecule Set</returns>
/// <exception cref="CDKException"></exception>
private static IChemObjectSet <IAtomContainer> GetUncommon(IAtomContainer mol, IAtomContainer mcss, bool shouldMatchBonds)
{
List <int> atomSerialsToDelete = new List <int>();
var matches = CDKMCS.GetSubgraphAtomsMaps(mol, mcss, shouldMatchBonds);
var mapList = matches[0];
foreach (var o in mapList)
{
CDKRMap rmap = (CDKRMap)o;
atomSerialsToDelete.Add(rmap.Id1);
}
// at this point we have the serial numbers of the bonds to delete
// we should get the actual bonds rather than delete by serial numbers
List <IAtom> atomsToDelete = new List <IAtom>();
foreach (var serial in atomSerialsToDelete)
{
atomsToDelete.Add(mol.Atoms[serial]);
}
// now lets get rid of the bonds themselves
foreach (var atom in atomsToDelete)
{
mol.RemoveAtomAndConnectedElectronContainers(atom);
}
// now we probably have a set of disconnected components
// so lets get a set of individual atom containers for
// corresponding to each component
return(ConnectivityChecker.PartitionIntoMolecules(mol));
}
public void TestTwentyRandomStructures()
{
var molecule = TestMoleculeFactory.MakeAlphaPinene();
var rg = new RandomGenerator(molecule);
IAtomContainer result = null;
for (int f = 0; f < 50; f++)
{
result = rg.ProposeStructure();
Assert.AreEqual(molecule.Atoms.Count, result.Atoms.Count);
Assert.AreEqual(1, ConnectivityChecker.PartitionIntoMolecules(result).Count());
}
}
/// <summary>
/// Fragment a molecule
/// </summary>
/// <param name="mol"></param>
/// <returns></returns>
public static List <KeyValuePair <string, IAtomContainer> > FragmentMoleculeAndCanonicalizeSmiles(
IAtomContainer mol,
bool filterOutCommonCounterIons)
{
int aci, fi, i1;
KeyValuePair <string, IAtomContainer> kvp;
List <KeyValuePair <string, IAtomContainer> > frags = new List <KeyValuePair <string, IAtomContainer> >();
if (mol == null)
{
return(frags);
}
IReadOnlyList <IAtomContainer> acs = ConnectivityChecker.PartitionIntoMolecules(mol);
int acc = acs.Count;
for (aci = 0; aci < acc; aci++)
{
IAtomContainer fragMol = acs[aci];
string fragSmiles = AtomContainerToSmiles(fragMol);
if (filterOutCommonCounterIons)
{
if (CommonSmallFragments.Contains(fragSmiles) ||
GetHeavyAtomCount(fragMol) <= 6)
{
continue;
}
}
kvp = new KeyValuePair <string, IAtomContainer>(fragSmiles, fragMol);
int ac = fragMol.Atoms.Count;
for (fi = frags.Count - 1; fi >= 0; fi--) // insert into list so that fragments are ordered largest to smallest
{
if (frags[fi].Value.Atoms.Count >= ac)
{
break;
}
}
frags.Insert(fi + 1, kvp);
}
return(frags);
}
/// <summary>
/// Modules for cleaning a molecule
/// </summary>
/// <param name="molecule"></param>
/// <returns>cleaned AtomContainer</returns>
public static IAtomContainer CheckAndCleanMolecule(IAtomContainer molecule)
{
bool isMarkush = false;
foreach (var atom in molecule.Atoms)
{
if (atom.Symbol.Equals("R", StringComparison.Ordinal))
{
isMarkush = true;
break;
}
}
if (isMarkush)
{
Console.Error.WriteLine("Skipping Markush structure for sanity check");
}
// Check for salts and such
if (!ConnectivityChecker.IsConnected(molecule))
{
// lets see if we have just two parts if so, we assume its a salt and just work
// on the larger part. Ideally we should have a check to ensure that the smaller
// part is a metal/halogen etc.
var fragments = ConnectivityChecker.PartitionIntoMolecules(molecule).ToReadOnlyList();
if (fragments.Count > 2)
{
Console.Error.WriteLine("More than 2 components. Skipped");
}
else
{
var frag1 = fragments[0];
var frag2 = fragments[1];
if (frag1.Atoms.Count > frag2.Atoms.Count)
{
molecule = frag1;
}
else
{
molecule = frag2;
}
}
}
Configure(molecule);
return(molecule);
}
public static void RemoveElectronContainer(IChemObjectSet <IAtomContainer> set, IElectronContainer electrons)
{
foreach (var atomContainer in set)
{
if (atomContainer.Contains(electrons))
{
atomContainer.Remove(electrons);
var molecules = ConnectivityChecker.PartitionIntoMolecules(atomContainer);
if (molecules.Count > 1)
{
set.Remove(atomContainer);
for (int k = 0; k < molecules.Count; k++)
{
set.Add(molecules[k]);
}
}
return;
}
}
}
private static int GetFragmentCount(IAtomContainer molecule)
{
bool fragmentFlag = true;
var fragmentMolSet = ChemObjectBuilder.Instance.NewAtomContainerSet();
int countFrag = 0;
if (molecule.Atoms.Count > 0)
{
fragmentFlag = ConnectivityChecker.IsConnected(molecule);
if (!fragmentFlag)
{
fragmentMolSet.AddRange(ConnectivityChecker.PartitionIntoMolecules(molecule));
}
else
{
fragmentMolSet.Add(molecule);
}
countFrag = fragmentMolSet.Count;
}
return(countFrag);
}
/// <summary>
/// Determine the ring set for this atom container.
/// </summary>
/// <param name="atomContainer">the atom container to find rings in.</param>
/// <returns>the rings of the molecule</returns>
protected virtual IRingSet GetRingSet(IAtomContainer atomContainer)
{
IRingSet ringSet = atomContainer.Builder.NewRingSet();
try
{
var molecules = ConnectivityChecker.PartitionIntoMolecules(atomContainer);
foreach (var mol in molecules)
{
ringSet.Add(Cycles.FindSSSR(mol).ToRingSet());
}
return(ringSet);
}
catch (Exception exception)
{
Trace.TraceWarning("Could not partition molecule: " + exception.Message);
Debug.WriteLine(exception);
return(ringSet);
}
}
/// <summary>
/// Generates a shortest path based BitArray fingerprint for the given AtomContainer.
/// </summary>
/// <param name="ac">The AtomContainer for which a fingerprint is generated</param>
/// <exception cref="CDKException">if there error in aromaticity perception or other CDK functions</exception>
/// <returns>A <see cref="BitArray"/> representing the fingerprint</returns>
public override IBitFingerprint GetBitFingerprint(IAtomContainer ac)
{
IAtomContainer atomContainer = null;
atomContainer = (IAtomContainer)ac.Clone();
Aromaticity.CDKLegacy.Apply(atomContainer);
var bitSet = new BitArray(fingerprintLength);
if (!ConnectivityChecker.IsConnected(atomContainer))
{
var partitionedMolecules = ConnectivityChecker.PartitionIntoMolecules(atomContainer);
foreach (var container in partitionedMolecules)
{
AddUniquePath(container, bitSet);
}
}
else
{
AddUniquePath(atomContainer, bitSet);
}
return(new BitSetFingerprint(bitSet));
}
/// <summary>
/// Initiates the process for the given mechanism. The atoms to apply are mapped between
/// reactants and products.
/// </summary>
/// <param name="atomContainerSet"></param>
/// <param name="atomList">
/// The list of atoms taking part in the mechanism. Only allowed two atoms.
/// The first atom is the atom which contains the ISingleElectron and the second
/// third is the atom which will be removed
/// the first atom</param>
/// <param name="bondList">The list of bonds taking part in the mechanism. Only allowed one bond.
/// It is the bond which is moved</param>
/// <returns>The Reaction mechanism</returns>
public IReaction Initiate(IChemObjectSet <IAtomContainer> atomContainerSet, IList <IAtom> atomList, IList <IBond> bondList)
{
var atMatcher = CDK.AtomTypeMatcher;
if (atomContainerSet.Count != 1)
{
throw new CDKException("RadicalSiteIonizationMechanism only expects one IAtomContainer");
}
if (atomList.Count != 3)
{
throw new CDKException("RadicalSiteIonizationMechanism expects three atoms in the List");
}
if (bondList.Count != 2)
{
throw new CDKException("RadicalSiteIonizationMechanism only expect one bond in the List");
}
IAtomContainer molecule = atomContainerSet[0];
IAtomContainer reactantCloned;
reactantCloned = (IAtomContainer)molecule.Clone();
IAtom atom1 = atomList[0]; // Atom containing the ISingleElectron
IAtom atom1C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom1)];
IAtom atom2 = atomList[1]; // Atom
IAtom atom2C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom2)];
IAtom atom3 = atomList[2]; // Atom to be saved
IAtom atom3C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom3)];
IBond bond1 = bondList[0]; // Bond to increase the order
int posBond1 = molecule.Bonds.IndexOf(bond1);
IBond bond2 = bondList[1]; // Bond to remove
int posBond2 = molecule.Bonds.IndexOf(bond2);
BondManipulator.IncreaseBondOrder(reactantCloned.Bonds[posBond1]);
reactantCloned.Bonds.Remove(reactantCloned.Bonds[posBond2]);
var selectron = reactantCloned.GetConnectedSingleElectrons(atom1C);
reactantCloned.SingleElectrons.Remove(selectron.Last());
atom1C.Hybridization = Hybridization.Unset;
AtomContainerManipulator.PercieveAtomTypesAndConfigureAtoms(reactantCloned);
IAtomType type = atMatcher.FindMatchingAtomType(reactantCloned, atom1C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
atom2C.Hybridization = Hybridization.Unset;
AtomContainerManipulator.PercieveAtomTypesAndConfigureAtoms(reactantCloned);
type = atMatcher.FindMatchingAtomType(reactantCloned, atom2C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
reactantCloned.SingleElectrons.Add(atom2C.Builder.NewSingleElectron(atom3C));
atom3C.Hybridization = Hybridization.Unset;
AtomContainerManipulator.PercieveAtomTypesAndConfigureAtoms(reactantCloned);
type = atMatcher.FindMatchingAtomType(reactantCloned, atom3C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
IReaction reaction = atom2C.Builder.NewReaction();
reaction.Reactants.Add(molecule);
/* mapping */
foreach (var atom in molecule.Atoms)
{
IMapping mapping = atom2C.Builder.NewMapping(atom,
reactantCloned.Atoms[molecule.Atoms.IndexOf(atom)]);
reaction.Mappings.Add(mapping);
}
var moleculeSetP = ConnectivityChecker.PartitionIntoMolecules(reactantCloned);
foreach (var moleculeP in moleculeSetP)
{
reaction.Products.Add(moleculeP);
}
return(reaction);
}
/// <summary>
/// Initiates the process for the given mechanism. The atoms to apply are mapped between
/// reactants and products.
/// </summary>
/// <param name="atomContainerSet"></param>
/// <param name="atomList">The list of atoms taking part in the mechanism. Only allowed two atoms. The first atom receives the charge and the second the single electron</param>
/// <param name="bondList">The list of bonds taking part in the mechanism. Only allowed one bond</param>
/// <returns>The Reaction mechanism</returns>
public IReaction Initiate(IChemObjectSet <IAtomContainer> atomContainerSet, IList <IAtom> atomList, IList <IBond> bondList)
{
var atMatcher = CDKAtomTypeMatcher.GetInstance(CDKAtomTypeMatcher.Mode.RequireExplicitHydrogens);
if (atomContainerSet.Count != 1)
{
throw new CDKException("RemovingSEofBMechanism only expects one IAtomContainer");
}
if (atomList.Count != 2)
{
throw new CDKException("RemovingSEofBMechanism expects two atoms in the List");
}
if (bondList.Count != 1)
{
throw new CDKException("RemovingSEofBMechanism only expects one bond in the List");
}
var molecule = atomContainerSet[0];
var reactantCloned = (IAtomContainer)molecule.Clone();
var atom1 = atomList[0];
var atom1C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom1)];
var atom2 = atomList[1];
var atom2C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom2)];
var bond1 = bondList[0];
var posBond1 = molecule.Bonds.IndexOf(bond1);
if (bond1.Order == BondOrder.Single)
{
reactantCloned.Bonds.Remove(reactantCloned.Bonds[posBond1]);
}
else
{
BondManipulator.DecreaseBondOrder(reactantCloned.Bonds[posBond1]);
}
var charge = atom1C.FormalCharge.Value;
atom1C.FormalCharge = charge + 1;
reactantCloned.SingleElectrons.Add(atom1C.Builder.NewSingleElectron(atom2C));
// check if resulting atom type is reasonable
atom1C.Hybridization = Hybridization.Unset;
AtomContainerManipulator.PercieveAtomTypesAndConfigureAtoms(reactantCloned);
var type = atMatcher.FindMatchingAtomType(reactantCloned, atom1C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
atom2C.Hybridization = Hybridization.Unset;
AtomContainerManipulator.PercieveAtomTypesAndConfigureAtoms(reactantCloned);
type = atMatcher.FindMatchingAtomType(reactantCloned, atom2C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
var reaction = atom1C.Builder.NewReaction();
reaction.Reactants.Add(molecule);
/* mapping */
foreach (var atom in molecule.Atoms)
{
IMapping mapping = atom1C.Builder.NewMapping(atom,
reactantCloned.Atoms[molecule.Atoms.IndexOf(atom)]);
reaction.Mappings.Add(mapping);
}
if (bond1.Order != BondOrder.Single)
{
reaction.Products.Add(reactantCloned);
}
else
{
var moleculeSetP = ConnectivityChecker.PartitionIntoMolecules(reactantCloned);
foreach (var moleculeP in moleculeSetP)
{
reaction.Products.Add(moleculeP);
}
}
return(reaction);
}
/// <summary>
/// Find all enabled abbreviations in the provided molecule. They are not
/// added to the existing Sgroups and may need filtering.
/// </summary>
/// <param name="mol">molecule</param>
/// <returns>list of new abbreviation Sgroups</returns>
public IList <Sgroup> Generate(IAtomContainer mol)
{
// mark which atoms have already been abbreviated or are
// part of an existing Sgroup
var usedAtoms = new HashSet <IAtom>();
var sgroups = mol.GetCtabSgroups();
if (sgroups != null)
{
foreach (var sgroup in sgroups)
{
foreach (var atom in sgroup.Atoms)
{
usedAtoms.Add(atom);
}
}
}
var newSgroups = new List <Sgroup>();
// disconnected abbreviations, salts, common reagents, large compounds
if (!usedAtoms.Any())
{
try
{
var copy = AtomContainerManipulator.CopyAndSuppressedHydrogens(mol);
string cansmi = usmigen.Create(copy);
if (disconnectedAbbreviations.TryGetValue(cansmi, out string label) && !disabled.Contains(label) && ContractToSingleLabel)
{
var sgroup = new Sgroup
{
Type = SgroupType.CtabAbbreviation,
Subscript = label
};
foreach (var atom in mol.Atoms)
{
sgroup.Atoms.Add(atom);
}
return(new[] { sgroup });
}
else if (cansmi.Contains("."))
{
var parts = ConnectivityChecker.PartitionIntoMolecules(mol);
// leave one out
Sgroup best = null;
for (int i = 0; i < parts.Count; i++)
{
var a = parts[i];
var b = a.Builder.NewAtomContainer();
for (int j = 0; j < parts.Count; j++)
{
if (j != i)
{
b.Add(parts[j]);
}
}
var sgroup1 = GetAbbr(a);
var sgroup2 = GetAbbr(b);
if (sgroup1 != null && sgroup2 != null && ContractToSingleLabel)
{
var combined = new Sgroup();
label = null;
foreach (var atom in sgroup1.Atoms)
{
combined.Atoms.Add(atom);
}
foreach (var atom in sgroup2.Atoms)
{
combined.Atoms.Add(atom);
}
if (sgroup1.Subscript.Length > sgroup2.Subscript.Length)
{
combined.Subscript = sgroup1.Subscript + String_Interpunct + sgroup2.Subscript;
}
else
{
combined.Subscript = sgroup2.Subscript + String_Interpunct + sgroup1.Subscript;
}
combined.Type = SgroupType.CtabAbbreviation;
return(new[] { combined });
}
if (sgroup1 != null && (best == null || sgroup1.Atoms.Count > best.Atoms.Count))
{
best = sgroup1;
}
if (sgroup2 != null && (best == null || sgroup2.Atoms.Count < best.Atoms.Count))
{
best = sgroup2;
}
}
if (best != null)
{
newSgroups.Add(best);
foreach (var atom in best.Atoms)
{
usedAtoms.Add(atom);
}
}
}
}
catch (CDKException)
{
}
}
var fragments = GenerateFragments(mol);
var sgroupAdjs = new MultiDictionary <IAtom, Sgroup>();
foreach (var frag in fragments)
{
try
{
var smi = usmigen.Create(AtomContainerManipulator.CopyAndSuppressedHydrogens(frag));
if (!connectedAbbreviations.TryGetValue(smi, out string label) || disabled.Contains(label))
{
continue;
}
bool overlap = false;
// note: first atom is '*'
var numAtoms = frag.Atoms.Count;
var numBonds = frag.Bonds.Count;
for (int i = 1; i < numAtoms; i++)
{
if (usedAtoms.Contains(frag.Atoms[i]))
{
overlap = true;
break;
}
}
// overlaps with previous assignment
if (overlap)
{
continue;
}
// create new abbreviation Sgroup
var sgroup = new Sgroup
{
Type = SgroupType.CtabAbbreviation,
Subscript = label
};
var attachBond = frag.Bonds[0].GetProperty <IBond>(PropertyName_CutBond);
IAtom attachAtom = null;
sgroup.Bonds.Add(attachBond);
for (int i = 1; i < numAtoms; i++)
{
var atom = frag.Atoms[i];
usedAtoms.Add(atom);
sgroup.Atoms.Add(atom);
if (attachBond.Begin.Equals(atom))
{
attachAtom = attachBond.End;
}
else if (attachBond.End.Equals(atom))
{
attachAtom = attachBond.Begin;
}
}
if (attachAtom != null)
{
sgroupAdjs.Add(attachAtom, sgroup);
}
newSgroups.Add(sgroup);
}
catch (CDKException)
{
// ignore
}
}
if (!ContractOnHetero)
{
return(newSgroups);
}
// now collapse
foreach (var attach in mol.Atoms)
{
if (usedAtoms.Contains(attach))
{
continue;
}
// skip charged or isotopic labelled, C or R/*, H, He
if ((attach.FormalCharge != null && attach.FormalCharge != 0) ||
attach.MassNumber != null ||
attach.AtomicNumber == 6 ||
attach.AtomicNumber < 2)
{
continue;
}
var hcount = attach.ImplicitHydrogenCount.Value;
var xatoms = new HashSet <IAtom>();
var xbonds = new HashSet <IBond>();
var newbonds = new HashSet <IBond>();
xatoms.Add(attach);
var nbrSymbols = new List <string>();
var todelete = new HashSet <Sgroup>();
foreach (var sgroup in sgroupAdjs[attach])
{
if (ContainsChargeChar(sgroup.Subscript))
{
continue;
}
if (sgroup.Bonds.Count != 1)
{
continue;
}
var xbond = sgroup.Bonds.First();
xbonds.Add(xbond);
foreach (var a in sgroup.Atoms)
{
xatoms.Add(a);
}
if (attach.Symbol.Length == 1 &&
char.IsLower(sgroup.Subscript[0]))
{
if (ChemicalElement.OfSymbol(attach.Symbol + sgroup.Subscript[0]) != ChemicalElement.R)
{
goto continue_collapse;
}
}
nbrSymbols.Add(sgroup.Subscript);
todelete.Add(sgroup);
}
int numSGrpNbrs = nbrSymbols.Count;
foreach (var bond in mol.GetConnectedBonds(attach))
{
if (!xbonds.Contains(bond))
{
var nbr = bond.GetOther(attach);
// contract terminal bonds
if (mol.GetConnectedBonds(nbr).Count() == 1)
{
if (nbr.MassNumber != null ||
(nbr.FormalCharge != null && nbr.FormalCharge != 0))
{
newbonds.Add(bond);
}
else if (nbr.AtomicNumber == 1)
{
hcount++;
xatoms.Add(nbr);
}
else if (nbr.AtomicNumber > 0)
{
nbrSymbols.Add(NewSymbol(nbr.AtomicNumber, nbr.ImplicitHydrogenCount.Value, false));
xatoms.Add(nbr);
}
}
else
{
newbonds.Add(bond);
}
}
}
// reject if no symbols
// reject if no bonds (<1), except if all symbols are identical... (HashSet.size==1)
// reject if more that 2 bonds
if (!nbrSymbols.Any() ||
newbonds.Count < 1 && (new HashSet <string>(nbrSymbols).Count != 1) ||
newbonds.Count > 2)
{
continue;
}
// create the symbol
var sb = new StringBuilder();
sb.Append(NewSymbol(attach.AtomicNumber, hcount, newbonds.Count == 0));
string prev = null;
int count = 0;
nbrSymbols.Sort((o1, o2) =>
{
int cmp = o1.Length.CompareTo(o2.Length);
if (cmp != 0)
{
return(cmp);
}
return(o1.CompareTo(o2));
});
foreach (string nbrSymbol in nbrSymbols)
{
if (nbrSymbol.Equals(prev))
{
count++;
}
else
{
bool useParen = count == 0 || CountUpper(prev) > 1 || (prev != null && nbrSymbol.StartsWith(prev));
AppendGroup(sb, prev, count, useParen);
prev = nbrSymbol;
count = 1;
}
}
AppendGroup(sb, prev, count, false);
// remove existing
foreach (var e in todelete)
{
newSgroups.Remove(e);
}
// create new
var newSgroup = new Sgroup
{
Type = SgroupType.CtabAbbreviation,
Subscript = sb.ToString()
};
foreach (var bond in newbonds)
{
newSgroup.Bonds.Add(bond);
}
foreach (var atom in xatoms)
{
newSgroup.Atoms.Add(atom);
}
newSgroups.Add(newSgroup);
foreach (var a in xatoms)
{
usedAtoms.Add(a);
}
continue_collapse:
;
}
return(newSgroups);
}
/// <summary>
/// Performs the n point crossover of two <see cref="IAtomContainer"/>.
/// </summary>
/// <remarks>
/// Precondition: The atoms in the molecules are ordered by properties to
/// preserve (e. g. atom symbol). Due to its randomized nature, this method
/// fails in around 3% of all cases. A <see cref="CDKException"/> with message "Could not
/// mate these properly" will then be thrown.
/// </remarks>
/// <returns>The children.</returns>
/// <exception cref="CDKException">if it was not possible to form off springs.</exception>
public IReadOnlyList <IAtomContainer> DoCrossover(IAtomContainer dad, IAtomContainer mom)
{
int tries = 0;
while (true)
{
int dim = dad.Atoms.Count;
var redChild = new IAtomContainer[2];
var blueChild = new IAtomContainer[2];
var redAtoms = new List <int>();
var blueAtoms = new List <int>();
// randomly divide atoms into two parts: redAtoms and blueAtoms.
if (splitMode == SplitMode.Random)
{
// better way to randomly divide atoms into two parts: redAtoms
// and blueAtoms.
for (int i = 0; i < dim; i++)
{
redAtoms.Add(i);
}
for (int i = 0; i < (dim - numatoms); i++)
{
int ranInt = RandomNumbersTool.RandomInt(0, redAtoms.Count - 1);
redAtoms.RemoveAt(ranInt);
blueAtoms.Add(ranInt);
}
}
else
{
// split graph using depth/breadth first traverse
var graph = new ChemGraph(dad)
{
NumAtoms = numatoms
};
if (splitMode == SplitMode.DepthFirst)
{
redAtoms = graph.PickDFGraph().ToList();
}
else
{
//this is SPLIT_MODE_BREADTH_FIRST
redAtoms = graph.PickBFGraph().ToList();
}
for (int i = 0; i < dim; i++)
{
int element = i;
if (!(redAtoms.Contains(element)))
{
blueAtoms.Add(element);
}
}
}
/* * dividing over ** */
redChild[0] = dad.Builder.NewAtomContainer(dad);
blueChild[0] = dad.Builder.NewAtomContainer(dad);
redChild[1] = dad.Builder.NewAtomContainer(mom);
blueChild[1] = dad.Builder.NewAtomContainer(mom);
var blueAtomsInRedChild0 = new List <IAtom>();
for (int j = 0; j < blueAtoms.Count; j++)
{
blueAtomsInRedChild0.Add(redChild[0].Atoms[(int)blueAtoms[j]]);
}
for (int j = 0; j < blueAtomsInRedChild0.Count; j++)
{
redChild[0].RemoveAtom(blueAtomsInRedChild0[j]);
}
var blueAtomsInRedChild1 = new List <IAtom>();
for (int j = 0; j < blueAtoms.Count; j++)
{
blueAtomsInRedChild1.Add(redChild[1].Atoms[(int)blueAtoms[j]]);
}
for (int j = 0; j < blueAtomsInRedChild1.Count; j++)
{
redChild[1].RemoveAtom(blueAtomsInRedChild1[j]);
}
var redAtomsInBlueChild0 = new List <IAtom>();
for (int j = 0; j < redAtoms.Count; j++)
{
redAtomsInBlueChild0.Add(blueChild[0].Atoms[(int)redAtoms[j]]);
}
for (int j = 0; j < redAtomsInBlueChild0.Count; j++)
{
blueChild[0].RemoveAtom(redAtomsInBlueChild0[j]);
}
var redAtomsInBlueChild1 = new List <IAtom>();
for (int j = 0; j < redAtoms.Count; j++)
{
redAtomsInBlueChild1.Add(blueChild[1].Atoms[(int)redAtoms[j]]);
}
for (int j = 0; j < redAtomsInBlueChild1.Count; j++)
{
blueChild[1].RemoveAtom(redAtomsInBlueChild1[j]);
}
//if the two fragments of one and only one parent have an uneven number
//of attachment points, we need to rearrange them
var satCheck = CDK.SaturationChecker;
double red1attachpoints = 0;
for (int i = 0; i < redChild[0].Atoms.Count; i++)
{
red1attachpoints += satCheck.GetCurrentMaxBondOrder(redChild[0].Atoms[i], redChild[0]);
}
double red2attachpoints = 0;
for (int i = 0; i < redChild[1].Atoms.Count; i++)
{
red2attachpoints += satCheck.GetCurrentMaxBondOrder(redChild[1].Atoms[i], redChild[1]);
}
bool isok = true;
if (red1attachpoints % 2 == 1 ^ red2attachpoints % 2 == 1)
{
isok = false;
var firstToBalance = redChild[1];
var secondToBalance = blueChild[0];
if (red1attachpoints % 2 == 1)
{
firstToBalance = redChild[0];
secondToBalance = blueChild[1];
}
//we need an atom which has
//- an uneven number of "attachment points" and
//- an even number of outgoing bonds
foreach (var atom in firstToBalance.Atoms)
{
if (satCheck.GetCurrentMaxBondOrder(atom, firstToBalance) % 2 == 1 &&
firstToBalance.GetBondOrderSum(atom) % 2 == 0)
{
//we remove this from it's current container and add it to the other one
firstToBalance.RemoveAtom(atom);
secondToBalance.Atoms.Add(atom);
isok = true;
break;
}
}
}
//if we have combinable fragments
if (isok)
{
//combine the fragments crosswise
var newstrucs = new IChemObjectSet <IAtomContainer> [2];
newstrucs[0] = dad.Builder.NewAtomContainerSet();
newstrucs[0].AddRange(ConnectivityChecker.PartitionIntoMolecules(redChild[0]));
newstrucs[0].AddRange(ConnectivityChecker.PartitionIntoMolecules(blueChild[1]));
newstrucs[1] = dad.Builder.NewAtomContainerSet();
newstrucs[1].AddRange(ConnectivityChecker.PartitionIntoMolecules(redChild[1]));
newstrucs[1].AddRange(ConnectivityChecker.PartitionIntoMolecules(blueChild[0]));
//and merge
var children = new List <IAtomContainer>(2);
for (int f = 0; f < 2; f++)
{
var structrue = pfsm.Generate2(newstrucs[f]);
if (structrue != null)
{
//if children are not correct, the outer loop will repeat,
//so we ignore this
children.Add(structrue);
}
}
if (children.Count == 2 &&
ConnectivityChecker.IsConnected(children[0]) &&
ConnectivityChecker.IsConnected(children[1]))
{
return(children);
}
}
tries++;
if (tries > 20)
{
throw new CDKException("Could not mate these properly");
}
}
}
/// <summary>
/// Initiates the process for the given mechanism. The atoms to apply are mapped between
/// reactants and products.
/// </summary>
/// <param name="atomContainerSet"></param>
/// <param name="atomList">The list of atoms taking part in the mechanism. Only allowed two three.
/// The first atom is the atom which must contain the charge to be moved, the second
/// is the atom which is in the middle and the third is the atom which acquires the new charge
/// <param name="bondList">The list of bonds taking part in the mechanism. Only allowed two bond.</param>
/// The first bond is the bond to increase the order and the second is the bond
/// to decrease the order
/// It is the bond which is moved</param>
/// <returns>The Reaction mechanism</returns>
public IReaction Initiate(IChemObjectSet <IAtomContainer> atomContainerSet, IList <IAtom> atomList, IList <IBond> bondList)
{
var atMatcher = CDK.AtomTypeMatcher;
if (atomContainerSet.Count != 1)
{
throw new CDKException("RearrangementChargeMechanism only expects one IAtomContainer");
}
if (atomList.Count != 3)
{
throw new CDKException("RearrangementChargeMechanism expects three atoms in the List");
}
if (bondList.Count != 2)
{
throw new CDKException("RearrangementChargeMechanism only expect one bond in the List");
}
IAtomContainer molecule = atomContainerSet[0];
IAtomContainer reactantCloned;
reactantCloned = (IAtomContainer)molecule.Clone();
IAtom atom1 = atomList[0]; // Atom with the charge
IAtom atom1C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom1)];
IAtom atom3 = atomList[2]; // Atom which acquires the charge
IAtom atom3C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom3)];
IBond bond1 = bondList[0]; // Bond with single bond
int posBond1 = molecule.Bonds.IndexOf(bond1);
IBond bond2 = bondList[1]; // Bond with double bond
int posBond2 = molecule.Bonds.IndexOf(bond2);
BondManipulator.IncreaseBondOrder(reactantCloned.Bonds[posBond1]);
if (bond2.Order == BondOrder.Single)
{
reactantCloned.Bonds.Remove(reactantCloned.Bonds[posBond2]);
}
else
{
BondManipulator.DecreaseBondOrder(reactantCloned.Bonds[posBond2]);
}
//Depending of the charge moving (radical, + or -) there is a different situation
if (reactantCloned.GetConnectedSingleElectrons(atom1C).Any())
{
var selectron = reactantCloned.GetConnectedSingleElectrons(atom1C);
reactantCloned.SingleElectrons.Remove(selectron.Last());
reactantCloned.SingleElectrons.Add(bond2.Builder.NewSingleElectron(atom3C));
}
else if (atom1C.FormalCharge > 0)
{
int charge = atom1C.FormalCharge.Value;
atom1C.FormalCharge = charge - 1;
charge = atom3C.FormalCharge.Value;
atom3C.FormalCharge = charge + 1;
}
else if (atom1C.FormalCharge < 1)
{
int charge = atom1C.FormalCharge.Value;
atom1C.FormalCharge = charge + 1;
var ln = reactantCloned.GetConnectedLonePairs(atom1C);
reactantCloned.LonePairs.Remove(ln.Last());
atom1C.IsAromatic = false;
charge = atom3C.FormalCharge.Value;
atom3C.FormalCharge = charge - 1;
reactantCloned.LonePairs.Add(bond2.Builder.NewLonePair(atom3C));
atom3C.IsAromatic = false;
}
else
{
return(null);
}
atom1C.Hybridization = Hybridization.Unset;
atom3C.Hybridization = Hybridization.Unset;
AtomContainerManipulator.PercieveAtomTypesAndConfigureAtoms(reactantCloned);
IAtomType type = atMatcher.FindMatchingAtomType(reactantCloned, atom1C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
type = atMatcher.FindMatchingAtomType(reactantCloned, atom3C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
IReaction reaction = bond2.Builder.NewReaction();
reaction.Reactants.Add(molecule);
/* mapping */
foreach (var atom in molecule.Atoms)
{
IMapping mapping = bond2.Builder.NewMapping(atom,
reactantCloned.Atoms[molecule.Atoms.IndexOf(atom)]);
reaction.Mappings.Add(mapping);
}
if (bond2.Order != BondOrder.Single)
{
reaction.Products.Add(reactantCloned);
}
else
{
IChemObjectSet <IAtomContainer> moleculeSetP = ConnectivityChecker.PartitionIntoMolecules(reactantCloned);
for (int z = 0; z < moleculeSetP.Count(); z++)
{
reaction.Products.Add((IAtomContainer)moleculeSetP[z]);
}
}
return(reaction);
}
/// <summary>
/// Parse a reaction SMILES.
/// </summary>
/// <param name="smiles">The SMILES string to parse</param>
/// <returns>An instance of <see cref="IReaction"/></returns>
/// <exception cref="InvalidSmilesException">if the string cannot be parsed</exception>
/// <seealso cref="ParseSmiles(string)"/>
public IReaction ParseReactionSmiles(string smiles)
{
if (!smiles.Contains(">"))
{
throw new InvalidSmilesException("Not a reaction SMILES: " + smiles);
}
var first = smiles.IndexOf('>');
var second = smiles.IndexOf('>', first + 1);
if (second < 0)
{
throw new InvalidSmilesException("Invalid reaction SMILES:" + smiles);
}
var reactants = smiles.Substring(0, first);
var agents = smiles.Substring(first + 1, second - (first + 1));
var products = smiles.Substring(second + 1, smiles.Length - (second + 1));
var reaction = builder.NewReaction();
// add reactants
if (!(reactants.Count() == 0))
{
var reactantContainer = ParseSmiles(reactants, true);
var reactantSet = ConnectivityChecker.PartitionIntoMolecules(reactantContainer);
foreach (var reactant in reactantSet)
{
reaction.Reactants.Add(reactant);
}
}
// add agents
if (!(agents.Count() == 0))
{
var agentContainer = ParseSmiles(agents, true);
var agentSet = ConnectivityChecker.PartitionIntoMolecules(agentContainer);
foreach (var agent in agentSet)
{
reaction.Agents.Add(agent);
}
}
string title = null;
// add products
if (!(products.Count() == 0))
{
var productContainer = ParseSmiles(products, true);
var productSet = ConnectivityChecker.PartitionIntoMolecules(productContainer);
foreach (var product in productSet)
{
reaction.Products.Add(product);
}
reaction.SetProperty(CDKPropertyName.Title, title = productContainer.Title);
}
try
{
//CXSMILES layer
ParseRxnCXSMILES(title, reaction);
}
catch (Exception e)
{
//e.StackTrace
throw new InvalidSmilesException("Error parsing CXSMILES:" + e.Message);
}
return(reaction);
}
/// <summary>
/// Initiates the process for the given mechanism. The atoms to apply are mapped between
/// reactants and products.
/// </summary>
/// <param name="atomContainerSet"></param>
/// <param name="atomList">The list of atoms taking part in the mechanism. Only allowed two atoms. Both atoms acquire a ISingleElectron</param>
/// <param name="bondList">The list of bonds taking part in the mechanism. Only allowed one bond</param>
/// <returns>The Reaction mechanism</returns>
public IReaction Initiate(IChemObjectSet <IAtomContainer> atomContainerSet, IList <IAtom> atomList, IList <IBond> bondList)
{
var atMatcher = CDK.AtomTypeMatcher;
if (atomContainerSet.Count != 1)
{
throw new CDKException("TautomerizationMechanism only expects one IAtomContainer");
}
if (atomList.Count != 2)
{
throw new CDKException("HomolyticCleavageMechanism expects two atoms in the List");
}
if (bondList.Count != 1)
{
throw new CDKException("HomolyticCleavageMechanism only expect one bond in the List");
}
IAtomContainer molecule = atomContainerSet[0];
IAtomContainer reactantCloned;
reactantCloned = (IAtomContainer)molecule.Clone();
IAtom atom1 = atomList[0];
IAtom atom1C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom1)];
IAtom atom2 = atomList[1];
IAtom atom2C = reactantCloned.Atoms[molecule.Atoms.IndexOf(atom2)];
IBond bond1 = bondList[0];
int posBond1 = molecule.Bonds.IndexOf(bond1);
if (bond1.Order == BondOrder.Single)
{
reactantCloned.Bonds.Remove(reactantCloned.Bonds[posBond1]);
}
else
{
BondManipulator.DecreaseBondOrder(reactantCloned.Bonds[posBond1]);
}
reactantCloned.SingleElectrons.Add(bond1.Builder.NewSingleElectron(atom1C));
reactantCloned.SingleElectrons.Add(bond1.Builder.NewSingleElectron(atom2C));
AtomContainerManipulator.PercieveAtomTypesAndConfigureAtoms(reactantCloned);
// check if resulting atom type is reasonable
atom1C.Hybridization = Hybridization.Unset;
IAtomType type = atMatcher.FindMatchingAtomType(reactantCloned, atom1C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
// check if resulting atom type is reasonable: an acceptor atom cannot be charged positive*/
atom2C.Hybridization = Hybridization.Unset;
type = atMatcher.FindMatchingAtomType(reactantCloned, atom2C);
if (type == null || type.AtomTypeName.Equals("X", StringComparison.Ordinal))
{
return(null);
}
IReaction reaction = atom2C.Builder.NewReaction();
reaction.Reactants.Add(molecule);
/* mapping */
foreach (var atom in molecule.Atoms)
{
IMapping mapping = atom2C.Builder.NewMapping(atom,
reactantCloned.Atoms[molecule.Atoms.IndexOf(atom)]);
reaction.Mappings.Add(mapping);
}
if (bond1.Order != BondOrder.Single)
{
reaction.Products.Add(reactantCloned);
}
else
{
var moleculeSetP = ConnectivityChecker.PartitionIntoMolecules(reactantCloned);
foreach (var moleculeP in moleculeSetP)
{
reaction.Products.Add(moleculeP);
}
}
return(reaction);
}
/// <summary>
/// BuildFingerprints
/// </summary>
public static void BuildFingerprints(
UniChemData icd)
{
IAtomContainer mol, mol2;
InChIGenerator ig = null;
int acc, aci, ci;
icd.Children.Clear();
string parentFIKHB = icd.GetFIKHB();
DateTime t0 = DateTime.Now;
mol = CdkMol.InChIToAtomContainer(icd.InChIString);
icd.CanonSmiles = CdkMol.AtomContainerToSmiles(mol);
InChIToAtomContainerTime += TimeOfDay.Delta(ref t0);
BitSetFingerprint fp = // generate a fingerprint
CdkFingerprint.BuildBitSetFingerprint(mol, FingerprintType.MACCS, -1, -1);
BuildFinterprintTime1 += TimeOfDay.Delta(ref t0);
icd.Fingerprint = fp;
if (ConnectivityChecker.IsConnected(mol))
{
return; // single fragment
}
InChIGeneratorFactory igf = InChIGeneratorFactory.Instance;
AtomContainerSet acs = (AtomContainerSet)ConnectivityChecker.PartitionIntoMolecules(mol);
PartitionIntoMoleculesTime += TimeOfDay.Delta(ref t0);
acc = acs.Count;
for (aci = 0; aci < acc; aci++)
{
mol2 = acs[aci];
GetAtomContainerTime += TimeOfDay.Delta(ref t0);
ig = igf.GetInChIGenerator(mol2);
if (!IsAcceptableInchiStatus(ig))
{
continue;
}
string childKey = ig.GetInChIKey();
string childFIKHB = UniChemUtil.GetFIKHB(childKey);
InChIGeneratorTime += TimeOfDay.Delta(ref t0);
fp = // generate a fingerprint for the fragment
CdkFingerprint.BuildBitSetFingerprint(mol2, FingerprintType.MACCS, -1, -1);
BuildFinterprintTime2 += TimeOfDay.Delta(ref t0);
for (ci = 0; ci < icd.Children.Count; ci++) // see if a dup child
{
if (icd.Children[ci].ChildFIKHB == childFIKHB)
{
break;
}
}
if (ci < icd.Children.Count)
{
continue; // skip if dup
}
UniChemFIKHBHierarchy fikhbHier = new UniChemFIKHBHierarchy();
fikhbHier.ParentFIKHB = parentFIKHB;
fikhbHier.ChildFIKHB = childFIKHB;
fikhbHier.CanonSmiles = CdkMol.AtomContainerToSmiles(mol2);
fikhbHier.Fingerprint = fp;
icd.Children.Add(fikhbHier);
}
return;
}