Molecule Class Reference

#include <molecule.h>

Inheritance diagram for Molecule:

DataContainer KCFMolecule List of all members.

Public Member Functions

Molecule construction functions
 Molecule ()
virtual ~Molecule ()
 Molecule (Molecule &aMolecule, bool bool_resetMorganIndex=true)
 Molecule (Molecule &molecule1, Molecule &molecule2, double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *))
 Molecule (Molecule &m1, Molecule &m2, double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(float, float, float), float edgeKernelParameter)
Molecule & operator= (const Molecule &aMolecule)
virtual Atom * addAtom (string aSymbol, bool resetSSSR=true) throw ( CError )
Atom * addAtom (Atom *, bool resetSSSR=true, bool resetMorganIndex=true) throw ( CError)
Bond * linkAtoms (Atom *aSource, Atom *aTarget, int aBondLabel, int aBondStereo=0, int aBondNotUsed=0, int aBondTopology=0, int aBondReactionCenter=0, bool resetSSSR=true)
Bond * linkAtoms (string aSource, string aTarget, int aBondLabel, int aBondStereo=0, int aBondNotUsed=0, int aBondTopology=0, int aBondReactionCenter=0, bool resetSSSR=true) throw ( CError )
Bond * linkAtoms (int firstAtom, int secondAtom, int aBondLabel, int aBondStereo=0, int aBondNotUsed=0, int aBondTopology=0, int aBondReactionCenter=0, bool resetSSSR=true) throw ( CError )
void readMOL (string aFilename, bool genericAtomType=false) throw (CError)
void readMOLOld (string aFilename) throw (CError)
void erase ()
void eraseHiddenAtoms ()
void eraseRings ()
void eraseAdjacency ()
void eraseWalks ()
Accessor functions
int getId () const
string getIdString ()
string getName ()
void setName (string aName)
void setActivity (float aNumber)
void setActivity (string aNumber)
float getActivity (bool silentMode=false) throw ( CError )
float hasActivity ()
void unsetActivity ()
vector< Atom * > & getAtoms ()
Atom * getAtom (int anId) throw ( CError )
Atom * getAtomByIndex (int ind)
bool atomExists (Atom *anAtom)
int numAtoms ()
int numHiddenAtoms ()
int numBonds ()
int numHiddenBonds ()
long bondSum ()
void unsetBondFlags ()
void unsetBondFlagsOriginal ()
bool isChiral ()
void select ()
void unSelect ()
bool isSelected ()
void setSortDescriptor (string aName, int aType)
int getSortDescriptorType ()
string getSortDescriptorName ()
Ring * getRingWithID (int anID, bool createIfMissing) throw ( CError )
bool hasRing (Ring *newRing, bool detectingRing=false) throw ( CError )
bool hasRing () throw ( CError )
int numRings () throw ( CError )
bool hasSSSRDetected ()
float getMW (bool silentError=false) throw ( CError )
int getNumAtoms ()
string getLocation ()
int getOriginalFormat ()
void setOriginalFormat (int a)
Molecule manipulation functions
int hideAtomsByIntDescriptor (string aDescriptorName, int aValue, bool refreshBonds=true)
void hideAtom (vector< Atom * >::iterator anAtomI)
void hideAtomAndToFromBonds (vector< Atom * >::iterator anAtomI)
void hideAtomAndToFromBonds (Atom *anAtom)
int hideHydrogens ()
bool isHiddenAtom (Atom *anAtom)
int restoreHiddenAtoms (bool flagRestoreBonds=true)
int restoreHiddenBonds ()
int refreshBonds ()
void eraseAtom (Atom *anAtom) throw ( CError )
void deleteBonds ()
void deleteHiddenAtoms ()
void setMorganLabels (int anOrder)
void setPerretLabels ()
int setUniqueMorganIndices ()
void resetMorganIndex ()
int getNumberOfDistinctMorganIndices (int anOrder)
int getMaxMorganIteration ()
int getNumCarbonsOfComponent (string aDescriptorName, int aValue)
int getNumNitrogensOfComponent (string aDescriptorName, int aValue)
int numAtomsNonCSkeleton ()
float atomicDistance (Atom *atom1, Atom *atom2)
void compute ()
void binClassifyFromDescriptor (string descriptorName, float value, bool smallerOrEqual=true)
void atomsLabelsListing (vector< string > *)
void atomsSymbolsListing (vector< string > *)
void bondsListing (vector< int > *)
void noTottersTransform ()
void threeDtransform (int nBins, double distMin, double distMax)
void readPartialCharges (string charges)
void setMorganChargesLabels (double threshold)
int hideSalts (stringstream *out)
int markFragments ()
void unmarkFragments ()
void hideAllFragmentsBut (int aFragmentNumber)
void writeFragments (ofstream *outStream)
int DFS (Atom *startAtom, string intDescriptorName, int markValue)
int detectSSSR ()
void resetSSSR ()
void setHasSSSR ()
Bond * checkEdges (Atom *)
Iterators
vector< Atom * >::iterator beginAtom ()
vector< Atom * >::iterator endAtom ()
map< Atom *, Bond * >::iterator beginBond (Atom *anAtom)
map< Atom *, Bond * >::iterator endBond (Atom *anAtom)
map< Atom *, Bond * >::iterator beginBond (int anId)
map< Atom *, Bond * >::iterator endBond (int anId)
map< int, int >::iterator beginComponentSizes ()
map< int, int >::iterator endComponentSizes ()
Marginalized and general kernel related functions
void setKashimaKernelProb (double aPq, bool skipSkeleton=false)
double sumPT ()
double sumProbabilities ()
double sumProbabilitiesFast ()
double sumPQPS ()
double sumPQPSFast ()
void raisePowerFast ()
double computeKernel (Molecule *anotherMolecule, double(*pt2GraphKernel)(Molecule *mol1, Molecule *mol2, double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int parameter1, int parameter2), double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int parameter1, int parameter2=1)
double getSelfKernel (double(*pt2GraphKernel)(Molecule *mol1, Molecule *mol2, double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int, int), double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int parameter1, int parameter2=1)
double calculateSelfKernel (double(*pt2GraphKernel)(Molecule *mol1, Molecule *mol2, double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int, int), double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int paramter1, int parameter2)
double getSelfKernel () throw ( CError )
void setSelfKernel (double value)
void addToSelfKernel (double)
void substractToSelfKernel (double)
void resetSelfKernel ()
Matrix utilities functions related to the 3D pharmacophore kernel (Mahe 2006)
void setAdjacency (int i, int j, double value)
double getAdjacency (int i, int j)
void setWalks (int i, int j, double value)
double getWalks (int i, int j)
void raisePowerAdjacency ()
double traceWalks ()
double traceDiagWalks ()
Output functions
string toString ()
string toStringShort ()
string toStringLong ()
void describe ()
void describeShort ()
void describeLong ()
void describeEachAtom ()
void writeMOL (string aFileName)
void writeDOT (string aFilename, bool perretLabels=false)
void writeSD (string aFileName)

Protected Member Functions

void moleculeChanged (bool resetSSSR=true, bool resetMorganIndex=true)

Protected Attributes

vector< Atom * > atoms
vector< Atom * > hiddenAtoms
vector< Ring * > sssr
bool flagHasSSSRDetected
int id
bool selectedFlag
double selfKernel
bool selfKernelCalculated
bool chiral
bool flagActivity
int sortDescriptorType
string sortDescriptorName
int maxMorganIteration

Static Protected Attributes

static int counter

Private Member Functions

Bond * linkAtomsNoReturn (Atom *aSource, Atom *aTarget, int aBondLabel, int aBondStereo=0, int aBondNotUsed=0, int aBondTopology=0, int aBondReactionCenter=0)
void linkAtomsNoReturn (int firstAtom, int secondAtom, int aBondLabel) throw ( CError )

Private Attributes

map< Atom *, map< Atom *,
double > * > * 		fastPT
map< Atom *, map< Atom *,
double > * > * 		fastPTNext
map< Atom *, map< Atom *,
double > * > * 		fastPTSave
map< Atom *, double > fastPQ
map< Atom *, double > fastPS
map< int, int > componentSizes
float activity
string location
int originalFormat
vector< vector< double > > * adjacency
vector< vector< double > > * walks

Detailed Description

Molecule class which can be compared using Graph Kernels

Author:
Dr Jean-Luc Perret (luc@kuicr.kyoto-u.ac.jp), Kyoto University, Japan
Version:
0.3
Date:
17 Jan 2004
CLASS NAME: Molecule

FOR: SNSF SPONSORED PROJECT

PURPOSE: This class implements the notion of molecule. Molecules have

Properties are implemented by deriving the Molecule from the DataContainer class which takes care of memory allocation for these descriptors. Therefore descriptors can be added and removed at runtime using the DataContainer functions.

The Molecule class can be seen as a graph containing a list of vertex (Atom), each vertex containing a list of edges (Bond) to the neighbour vertex (Atom). Edges are directed and have a label. Chemical bonds are not directed so bonds are always double, one in each direction. So a Molecule can be seen as a directed labeled graph with bidirectional edges. ... many other changes

Examples:

molecule_example.cpp.

Constructor & Destructor Documentation

Molecule::Molecule (   ) 
 

class constructor.

virtual Molecule::~Molecule (   )  [virtual]
 

class destructor. deletes all atoms in the molecule.

Molecule::Molecule (  Molecule &  aMolecule,
bool  bool_resetMorganIndex = true
) 
 

copy constructor.

Molecule::Molecule (  Molecule &  molecule1,
Molecule &  molecule2,
double(*)(Atom *, Atom *)  pt2AtomKernel,
double(*)(Bond *, Bond *)  pt2BondKernel
) 
 

product graph constructor. this constructor constructs a fusion molecule from two molecules using a product graph operation. two atoms which are connected in both molecules have an edge connecting them in the fused graph.

Molecule::Molecule (  Molecule &  m1,
Molecule &  m2,
double(*)(Atom *, Atom *)  pt2AtomKernel,
double(*)(float, float, float)  pt2BondKernel,
float  edgeKernelParameter
) 
 

"3D" constructor. Note : BondKernel prototype is different of those used on the "product graph constructor".

Member Function Documentation

Atom* Molecule::addAtom (  Atom *  ,
bool  resetSSSR = true,
bool  resetMorganIndex = true
)  throw ( CError)
 

adds an Atom to the atoms vector. Returns a pointer to the created Atom. WARNING this atom will be deleted when the molecule is deleted so don't use this pointer! if resetSSSR = true, then this action invalidates previous sssr computations.

virtual Atom* Molecule::addAtom (  string  aSymbol,
bool  resetSSSR = true
)  throw ( CError ) [virtual]
 

adds an Atom with aSymbol chemical symbol to the molecule. aSymbol can be in any case. The first letter of aSymbol will be automatically changed to capital and the second letter to normal caps. Returns a pointer to the created Atom. WARNING this atom will be deleted when the molecule is deleted so don't use this pointer! if resetSSSR = true, then this action invalidates previous sssr computations.

Examples:
molecule_example.cpp.

void Molecule::addToSelfKernel (  double   ) 
 

adds a value to the self kernel.

bool Molecule::atomExists (  Atom *  anAtom  ) 
 

checks if an atom is part of the molecule. returns true if anAtom is included in the atom vector, false otherwise.

float Molecule::atomicDistance (  Atom *  atom1,
Atom *  atom2
) 
 

computes the Euclidian distance between two atoms.

void Molecule::atomsLabelsListing (  vector< string > *   ) 
 

fills in a vector of String with the distinct labels of the atoms of the molecule.

void Molecule::atomsSymbolsListing (  vector< string > *   ) 
 

fills in vector of String with the distinct symbols of the atoms of the molecule.

vector<Atom*>::iterator Molecule::beginAtom (   )  [inline]
 

returns an iterator to the first Atom.

map<Atom*, Bond*>::iterator Molecule::beginBond (  int  anId  )  [inline]
 

returns an iterator to the first Bond of an Atom designated by its id.

map<Atom*, Bond*>::iterator Molecule::beginBond (  Atom *  anAtom  )  [inline]
 

returns an iterator to the first Bond of an Atom.

map<int, int>::iterator Molecule::beginComponentSizes (   )  [inline]
 

NOT DOCUMENTED

void Molecule::binClassifyFromDescriptor (  string  descriptorName,
float  value,
bool  smallerOrEqual = true
) 
 

sets the activity according to the value of a descriptor and a value. If true is given as third argument (default) then descriptorName <= value are considered positive. If false is given as third argument then molecules with descriptorName >= value are considered positive. if a descriptor is missing then the molecule is left without activity.

void Molecule::bondsListing (  vector< int > *   ) 
 

fills in vector of Int with the distinct types of bonds of the molecule.

long Molecule::bondSum (   ) 
 

returns the sum of the bond types.

double Molecule::calculateSelfKernel (  double(*)(Molecule *mol1, Molecule *mol2, double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int, int)  pt2GraphKernel,
double(*)(Atom *, Atom *)  pt2AtomKernel,
double(*)(Bond *, Bond *)  pt2BondKernel,
int  paramter1,
int  parameter2
) 
 

calculates the molecule's self Kashima kernel and stores the result in the private selfKernel variable.

Bond* Molecule::checkEdges (  Atom *   ) 
 

helper function for detectSSSR. selects an optimum edge for elimination in structures without N2 nodes.

void Molecule::compute (   ) 
 

this function should be called everytime a molecule is created / modified. automatically called at the end of Molecule::readMol() and MoleculeUtils::readMDLCtabBlock().

double Molecule::computeKernel (  Molecule *  anotherMolecule,
double(*)(Molecule *mol1, Molecule *mol2, double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int parameter1, int parameter2)  pt2GraphKernel,
double(*)(Atom *, Atom *)  pt2AtomKernel,
double(*)(Bond *, Bond *)  pt2BondKernel,
int  parameter1,
int  parameter2 = 1
) 
 

compares this molecule with anotherMolecule and returns the value defined by the function pt2GraphKernel, using the function pt2AtomKernel to compare atoms and pt2BondKernel to compare bonds.

void Molecule::deleteBonds (   ) 
 

deletes all bonds in all atoms.

void Molecule::deleteHiddenAtoms (   ) 
 

deletes all hidden atoms.

void Molecule::describe (   ) 
 

prints a description of the molecule to stdout using toString() and DataContainer describe().

Reimplemented from DataContainer.

void Molecule::describeEachAtom (   ) 
 

prints a description of each atom (using describe()) and each bond (using describe())to stdout.

void Molecule::describeLong (   ) 
 

prints a description of the molecule to stdout using the molecule toString(), all atoms toStringShort(), and all bonds toStringShort().

Examples:
molecule_example.cpp.

void Molecule::describeShort (   ) 
 

prints a description of the molecule to stdout using toString().

Reimplemented from DataContainer.

int Molecule::detectSSSR (   ) 
 

detects the smallest set of smallest rings and returns the number of rings. this procedure is automatically called when a molecule is loaded from a mol or sd file through the function compute. call compute() explicitly after modifying atom or bond composition. this function also sets the ring membership of atoms and bonds. according to: Ring Perception Using Breadth-First Search, John Figueras 399 Baker s Pond Road, Orleans, Massachusetts 02653, J. Chem. Inf. Comput. Sci. 1996, 36, 986-991.

int Molecule::DFS (  Atom *  startAtom,
string  intDescriptorName,
int  markValue
) 
 

performs a DFS search on the graph starting at startAtom, marking atoms with markValue in descriptor intDescriptorName, and returning the number of nodes in the connected component.

vector<Atom*>::iterator Molecule::endAtom (   )  [inline]
 

returns an iterator to the last Atom.

map<Atom*, Bond*>::iterator Molecule::endBond (  int  anId  )  [inline]
 

returns an iterator to the first Bond of an Atom designated by its id.

map<Atom*, Bond*>::iterator Molecule::endBond (  Atom *  anAtom  )  [inline]
 

returns an iterator to the first Bond of an Atom.

map<int, int>::iterator Molecule::endComponentSizes (   )  [inline]
 

NOT DOCUMENTED

void Molecule::erase (   ) 
 

erases all atoms in the current molecule. returns the molecule in the state of construction by the default constructor.

void Molecule::eraseAdjacency (   ) 
 

erases the adjacency matrix of the molecule.

void Molecule::eraseAtom (  Atom *  anAtom  )  throw ( CError )
 

erases anAtom from the list. Does not delete the atom though!

void Molecule::eraseHiddenAtoms (   ) 
 

erases hidden atoms of the molecule.

void Molecule::eraseRings (   ) 
 

erases the rings of the molecule.

void Molecule::eraseWalks (   ) 
 

erases the 'walks' matrix of the molecule (used to compute the pharmacophore kernel, see (Mahe, 2006)).

float Molecule::getActivity (  bool  silentMode = false  )  throw ( CError )
 

returns the activity of the molecule.

double Molecule::getAdjacency (  int  i,
int  j
) 
 

returns the value of an entry of the adjacency matrix of the molecule.

Atom* Molecule::getAtom (  int  anId  )  throw ( CError )
 

returns a pointer to an atom designated by its unique Id. raises a CError exception if no atom with this Id is present in the molecule.

vector<Atom*>& Molecule::getAtoms (   )  [inline]
 

returns a pointer to the atom's vector.

int Molecule::getId (   )  const [inline]
 

returns the unique id of the molecule.

string Molecule::getIdString (   ) 
 

returns the unique id of the molecule as a string.

string Molecule::getLocation (   )  [inline]
 

returns the location where the molecule is stored.

int Molecule::getMaxMorganIteration (   ) 
 

returns the iteration of the Morgan index at which the different connectivity values have reached a maximum.

float Molecule::getMW (  bool  silentError = false  )  throw ( CError )
 

returns the molecular weight.

string Molecule::getName (   )  [inline]
 

returns the molecule name, i.e., the name descriptor value (Nb :not necessarily unique).

int Molecule::getNumAtoms (   )  [inline]
 

returns the number of atoms in the molecule (ignoring hidden atoms).

int Molecule::getNumberOfDistinctMorganIndices (  int  anOrder  ) 
 

returns the number of distinct morgan indices (of order anOrder) present in the molecule.

int Molecule::getNumCarbonsOfComponent (  string  aDescriptorName,
int  aValue
) 
 

returns the number of carbons in a connected component graph defined with an Int descriptor named aDescriptorName having value aValue.

int Molecule::getNumNitrogensOfComponent (  string  aDescriptorName,
int  aValue
) 
 

returns the number of nitrogens in a connected component graph defined with an Int descriptor named aDescriptorName having value aValue.

int Molecule::getOriginalFormat (   )  [inline]
 

returns the format of the original file. one of MDL2000, MDL3000, MDL2000JLP.

Ring* Molecule::getRingWithID (  int  anID,
bool  createIfMissing
)  throw ( CError )
 

returns the ring of the corresponding Id.

double Molecule::getSelfKernel (   )  throw ( CError )
 

emits an error if the kernel was not calculated before.

double Molecule::getSelfKernel (  double(*)(Molecule *mol1, Molecule *mol2, double(*pt2AtomKernel)(Atom *, Atom *), double(*pt2BondKernel)(Bond *, Bond *), int, int)  pt2GraphKernel,
double(*)(Atom *, Atom *)  pt2AtomKernel,
double(*)(Bond *, Bond *)  pt2BondKernel,
int  parameter1,
int  parameter2 = 1
)  [inline]
 

returns the kernel value for the molecule with itself.

string Molecule::getSortDescriptorName (   )  [inline]
 

returns the name of the descriptor used to sort molecules.

int Molecule::getSortDescriptorType (   )  [inline]
 

returns the type of the descriptor used to sort molecules.

double Molecule::getWalks (  int  i,
int  j
) 
 

returns the value of an entry of the walks matrix of the molecule.

float Molecule::hasActivity (   )  [inline]
 

returns true if the activity of the molecule was initialized.

bool Molecule::hasRing (   )  throw ( CError )
 

returns true if the molecule has a ring. If sssr was not detected before this call, detectSSSR() is called.

bool Molecule::hasRing (  Ring *  newRing,
bool  detectingRing = false
)  throw ( CError )
 

returns true if the molecule already has newRing in sssr. If sssr was not detected before this call, detectSSSR() is called.

bool Molecule::hasSSSRDetected (   )  [inline]
 

returns the value of the 'hasSSRDetected' flag.

void Molecule::hideAllFragmentsBut (  int  aFragmentNumber  ) 
 

NOT DOCUMENTED

void Molecule::hideAtom (  vector< Atom * >::iterator  anAtomI  ) 
 

hides an atom (but not the bonds to this atom). call refreshBonds() after hiding all atoms.

void Molecule::hideAtomAndToFromBonds (  Atom *  anAtom  ) 
 

hides anAtom and all to / from bonds for this atom.

void Molecule::hideAtomAndToFromBonds (  vector< Atom * >::iterator  anAtomI  ) 
 

hides anAtom and all to / from bonds for this atom.

int Molecule::hideAtomsByIntDescriptor (  string  aDescriptorName,
int  aValue,
bool  refreshBonds = true
) 
 

hides all atoms (and associated edges) with intDescriptor aDescriptorName == aValue. returns the number of hidden atoms

int Molecule::hideHydrogens (   ) 
 

moves the hydrogen atoms to the hiddenAtoms container, so that further algorithms will run without hydrogens. returns the number of hydrogens hidden. hidden hydrogens can be restored using restoreHiddenAtoms().

int Molecule::hideSalts (  stringstream *  out  ) 
 

hides all but the biggest (in atom numbers) connected atoms in the molecule. returns the number of components removed. WARNING: does not test if two compounds have identical maximum size. in that case, it will hide all but the first one.

bool Molecule::isChiral (   )  [inline]
 

returns true if the molecule has chiral information.

bool Molecule::isHiddenAtom (  Atom *  anAtom  ) 
 

returns true if an atom is hidden.

bool Molecule::isSelected (   )  [inline]
 

returns selectedFlag.

Bond* Molecule::linkAtoms (  int  firstAtom,
int  secondAtom,
int  aBondLabel,
int  aBondStereo = 0,
int  aBondNotUsed = 0,
int  aBondTopology = 0,
int  aBondReactionCenter = 0,
bool  resetSSSR = true
)  throw ( CError )
 

links two atoms using their position in the atom container. if firstAtom or secondAtom are not between 0 and numAtoms()-1 a CError error is thrown. returns a pointer to the forward bond just created (the return bond can be retrived with forwardBond->getReverse()).

Bond* Molecule::linkAtoms (  string  aSource,
string  aTarget,
int  aBondLabel,
int  aBondStereo = 0,
int  aBondNotUsed = 0,
int  aBondTopology = 0,
int  aBondReactionCenter = 0,
bool  resetSSSR = true
)  throw ( CError )
 

links two atoms using their name. If no atom with name aSource or aTarget are found a CError error is thrown. returns a pointer to the forward bond just created (the return bond can be retrived with forwardBond->getReverse()).

Bond* Molecule::linkAtoms (  Atom *  aSource,
Atom *  aTarget,
int  aBondLabel,
int  aBondStereo = 0,
int  aBondNotUsed = 0,
int  aBondTopology = 0,
int  aBondReactionCenter = 0,
bool  resetSSSR = true
) 
 

links two atoms poited by aSource and aTarget using aBondLabel bond type. If the two atoms are not in the atoms vector they are first added to the molecule. returns a pointer to the forward bond just created (the return bond can be retrived with forwardBond->getReverse()).

Examples:
molecule_example.cpp.

void Molecule::linkAtomsNoReturn (  int  firstAtom,
int  secondAtom,
int  aBondLabel
)  throw ( CError ) [private]
 

links a pair of atoms in a single direction only : firstAtom -> secondAtom.

Bond* Molecule::linkAtomsNoReturn (  Atom *  aSource,
Atom *  aTarget,
int  aBondLabel,
int  aBondStereo = 0,
int  aBondNotUsed = 0,
int  aBondTopology = 0,
int  aBondReactionCenter = 0
)  [private]
 

links two atoms poited by aSource and aTarget using aBondLabel bond type. If the two atoms are not in the atoms vector they are first added to the molecule WARNING this function adds a single directed bond, only to be used with in Molecule::raisePower().

int Molecule::markFragments (   ) 
 

NOT DOCUMENTED

void Molecule::moleculeChanged (  bool  resetSSSR = true,
bool  resetMorganIndex = true
)  [protected]
 

this function is called internally every time a molecule is changed (atom or bonding), reset flags.

void Molecule::noTottersTransform (   ) 
 

transforms the molecule into a graph preventing 'totters' (see (Mahe 2004)).

int Molecule::numAtoms (   )  [inline]
 

returns the number of atoms of the molecule.

int Molecule::numAtomsNonCSkeleton (   ) 
 

returns the number of terminal atoms of the molecule.

int Molecule::numBonds (   ) 
 

returns the number of bonds of the molecule.

int Molecule::numHiddenAtoms (   )  [inline]
 

returns the number of hidden atoms of the molecule.

int Molecule::numHiddenBonds (   ) 
 

returns the number of hidden bonds of the molecule.

int Molecule::numRings (   )  throw ( CError ) [inline]
 

returns the number of rings present in the molecule.

Molecule& Molecule::operator= (  const Molecule &  aMolecule  ) 
 

assignment operator.

void Molecule::raisePowerAdjacency (   ) 
 

multiplies the walks matrix by the adjacency matrix.

void Molecule::raisePowerFast (   ) 
 

function used to compute the graph kernel using the fused graph + sum of the powers of the transition matrix approach. see MoleculeUtils::powerKernel(). same as raisePower() but instead of using the Bond class the transition probabilities are read from fastPt and fastPTSave.

void Molecule::readMOL (  string  aFilename,
bool  genericAtomType = false
)  throw (CError)
 

reads the molecule description from a Mol file. MUST STILL BE COMPLETED

void Molecule::readMOLOld (  string  aFilename  )  throw (CError)
 

OLD version of readMOL().

void Molecule::readPartialCharges (  string  charges  ) 
 

reads partial charges of the atoms of the molecule from a ';' separated string.

int Molecule::refreshBonds (   ) 
 

hides all bonds to and from hidden atoms. call this function after hiding all atoms in a molecule. returns the number of edges hidden.

void Molecule::resetMorganIndex (   ) 
 

resets the calculations of the morgan indices. CALL THIS FUNCTION EACH TIME THE MOLECULE IS MODIFIED!

void Molecule::resetSelfKernel (   )  [inline]
 

resets the self kernel.

void Molecule::resetSSSR (   )  [inline]
 

sets the flagHasSSSRDetected to false.

int Molecule::restoreHiddenAtoms (  bool  flagRestoreBonds = true  ) 
 

restores the hidden atoms (for example hydrogens).

int Molecule::restoreHiddenBonds (   ) 
 

restores the hidden bonds among atoms (does not restore bonds in hidden atoms).

void Molecule::select (   )  [inline]
 

sets selectedFlag.

void Molecule::setActivity (  string  aNumber  )  [inline]
 

sets the activity of the molecule from a string activity. provided for convenience.

void Molecule::setActivity (  float  aNumber  )  [inline]
 

sets the activity of the molecule (a descriptor in float native format for faster retrieval).

void Molecule::setAdjacency (  int  i,
int  j,
double  value
) 
 

sets an entry of the adjacency matrix of the molecule to a given value.

void Molecule::setHasSSSR (   )  [inline]
 

sets the flagHasSSSRDetected to true.

void Molecule::setKashimaKernelProb (  double  aPq,
bool  skipSkeleton = false
) 
 

sets the start, stop and transition probabilities for the calculation of the Kashima kernel as published in (Kashima et al. 2003). the start probability is set to an identical value of 1/numAtoms() for all atoms. the stop probability can be choosen between 0 and 1. the transition probability of each bond of an Atom is set to (1.0 - stop probability) / numbonds.

void Molecule::setMorganChargesLabels (  double  threshold  ) 
 

introduces the sign of the partial charge into the label of the atoms of the molecule.

void Molecule::setMorganLabels (  int  anOrder  ) 
 

sets the morganLabel of atoms to the concatenation of the atomic symbol and the Morgan index of anOrder iteration.

void Molecule::setName (  string  aName  )  [inline]
 

sets the molecule name.

Examples:
molecule_example.cpp.

void Molecule::setOriginalFormat (  int  a  )  [inline]
 

sets the molecule's original format. one of MDL2000, MDL3000, MDL2000JLP.

void Molecule::setPerretLabels (   ) 
 

sets the perretLabel of atoms: the atomic symbol. Only carbons get an additional J if they have more than two aromatic bonds (aromatic rings joining carbons).

void Molecule::setSelfKernel (  double  value  ) 
 

sets the set kernel to the given value.

void Molecule::setSortDescriptor (  string  aName,
int  aType
) 
 

sets the descriptor type and name used by the < operator to compare molecules. if reverse == true then the sorting will be in descending order.

int Molecule::setUniqueMorganIndices (   ) 
 

sets the uniqueMorganIndex of each atom to the Morgan index having the maximum of different connectivity values for the molecule. returns the iteration of the smallest Morgan Index with the maximum number of connectivity values.

void Molecule::setWalks (  int  i,
int  j,
double  value
) 
 

sets an entry of the 'walks' matrix (i.e., a weighted adjacency matrix) of the molecule to a given value.

void Molecule::substractToSelfKernel (  double   ) 
 

substracts a value to the self kernel.

double Molecule::sumPQPS (   ) 
 

function used to compute the graph kernel using the fused graph + sum of the powers of the transition matrix approach. see MoleculeUtils::powerKernel()

double Molecule::sumPQPSFast (   ) 
 

fast version of sumPQPS().

double Molecule::sumProbabilities (   ) 
 

function used to compute the graph kernel using the fused graph + sum the powers of the transition matrix approach. see MoleculeUtils::powerKernel().

double Molecule::sumProbabilitiesFast (   ) 
 

fast version of sumProbabilities().

double Molecule::sumPT (   ) 
 

returns the sum of transition probabilities. should be used with fused graph.

void Molecule::threeDtransform (  int  nBins,
double  distMin,
double  distMax
) 
 

transforms the molecule into a 3D graph (see (Mahe 2006)).

string Molecule::toString (   ) 
 

returns a string description of the molecule (name, unique Id, memory Adress, number of atoms).

string Molecule::toStringLong (   ) 
 

returns a string description of the molecule (name, unique Id, memory Adress, number of atoms and atom descriptions).

string Molecule::toStringShort (   ) 
 

returns a short string description of the molecule. (name and unique Id).

double Molecule::traceDiagWalks (   ) 
 

computes the diagonal elements of the walk matrix multiplied by the adjacency matrix, and sums to get the trace.

double Molecule::traceWalks (   ) 
 

computes the trace of the walks matrix.

void Molecule::unmarkFragments (   ) 
 

NOT DOCUMENTED

void Molecule::unSelect (   )  [inline]
 

unset selectedFlag.

void Molecule::unsetActivity (   )  [inline]
 

turns off the activity flag.

void Molecule::unsetBondFlags (   ) 
 

unsets all bonds flags.

void Molecule::unsetBondFlagsOriginal (   ) 
 

unsets all bonds flags original.

void Molecule::writeDOT (  string  aFilename,
bool  perretLabels = false
) 
 

writes a dot file representing the molecule. the molecule can be converted to graphic representation using Graphviz's dot program.

void Molecule::writeFragments (  ofstream *  outStream  ) 
 

NOT DOCUMENTED

void Molecule::writeMOL (  string  aFileName  ) 
 

write the molecule definition in a MDL MOL file.

void Molecule::writeSD (  string  aFileName  ) 
 

write a MDL structure data (SD) file with a single molecule.

Member Data Documentation

float Molecule::activity [private]
 

activity value of the molecule.

vector< vector<double> >* Molecule::adjacency [private]
 

adjacency matrix of the molecule. used to compute the 3D pharmacophore kernel.

vector<Atom*> Molecule::atoms [protected]
 

container for atoms.

bool Molecule::chiral [protected]
 

true if the molecule has chiral information.

map< int, int > Molecule::componentSizes [private]
 

NO DOCUMENTATION

int Molecule::counter [static, protected]
 

counter of the number of molecules which have been instanciated.

map<Atom*, double> Molecule::fastPQ [private]
 

hash used in the fast computations of the powers of the product graph adjacency matrix.

map<Atom*, double> Molecule::fastPS [private]
 

hash used in the fast computations of the powers of the product graph adjacency matrix.

map<Atom*, map<Atom*, double>* >* Molecule::fastPT [private]
 

hash used in the fast computations of the powers of the product graph adjacency matrix.

map<Atom*, map<Atom*, double>* >* Molecule::fastPTNext [private]
 

hash used in the fast computations of the powers of the product graph adjacency matrix.

map<Atom*, map<Atom*, double>* >* Molecule::fastPTSave [private]
 

hash used in the fast computations of the powers of the product graph adjacency matrix.

bool Molecule::flagActivity [protected]
 

true if the molecule has activity information.

bool Molecule::flagHasSSSRDetected [protected]
 

flag indicating if smallest set of smallest rings was detected.

vector<Atom*> Molecule::hiddenAtoms [protected]
 

container for hidden atoms (for example the removed hydrogens are moved to this container).

int Molecule::id [protected]
 

molecule unique Id.

string Molecule::location [private]
 

location where the molecule is stored. automatically set by readMOL.

int Molecule::maxMorganIteration [protected]
 

contains the iteration number at which the maximum number of distinct connectivity values is reached.

int Molecule::originalFormat [private]
 

original format of the molecule.

bool Molecule::selectedFlag [protected]
 

flag to select subsets of molecules.

double Molecule::selfKernel [protected]
 

molecule's self Kernel.

bool Molecule::selfKernelCalculated [protected]
 

true if selfKernel was calculated, false otherwise.

string Molecule::sortDescriptorName [protected]
 

name of the sorting descriptor.

int Molecule::sortDescriptorType [protected]
 

type and name of descriptor to be used to sort molecules. use setSortDescriptor( aType, aName ) getSortDescriptorType(), setSortDescriptorName() to access.

vector<Ring*> Molecule::sssr [protected]
 

container for rings (smallest set of smallest rings).

vector< vector<double> >* Molecule::walks [private]
 

'walks' matrix of the molecule. used to compute the 3D pharmacophore kernel.

The documentation for this class was generated from the following file:
Generated on Wed Nov 28 12:12:52 2007 for ChemCpp by  doxygen 1.4.6