00001 // 00002 // Copyright (C) 2004-2007 Greg Landrum and Rational Discovery LLC 00003 // 00004 // @@ All Rights Reserved @@ 00005 // 00006 00007 #ifndef _RD_EMBEDDER_H_ 00008 #define _RD_EMBEDDER_H_ 00009 00010 #include <GraphMol/ROMol.h> 00011 00012 namespace RDKit { 00013 namespace DGeomHelpers { 00014 00015 //! Compute an embedding (in 3D) for the specified molecule using Distance Geometry 00016 /*! 00017 The following operations are performed (in order) here: 00018 -# Build a distance bounds matrix based on the topology, including 1-5 00019 distances but not VDW scaling 00020 -# Triangle smooth this bounds matrix 00021 -# If step 2 fails - repeat step 1, this time without 1-5 bounds and with vdW 00022 scaling, and repeat step 2 00023 -# Pick a distance matrix at random using the bounds matrix 00024 -# Compute initial coordinates from the distance matrix 00025 -# Repeat steps 3 and 4 until maxIterations is reached or embedding is successful 00026 -# Adjust initial coordinates by minimizing a Distance Violation error function 00027 00028 \param mol Molecule of interest 00029 \param maxIterations Max. number of times the embedding will be tried if coordinates are 00030 not obtained successfully 00031 \param seed provides a seed for the random number generator (so that the same 00032 coordinates can be obtained for a molecule on multiple runs) 00033 If negative, the RNG will not be seeded. 00034 \param clearConfs Clear all existing conformations on the molecule 00035 \param useRandomCoords Start the embedding from random coordinates instead of 00036 using eigenvalues of the distance matrix. 00037 \param boxSizeMult Determines the size of the box that is used for 00038 random coordinates. If this is a positive number, the 00039 side length will equal the largest element of the distance 00040 matrix times \c boxSizeMult. If this is a negative number, 00041 the side length will equal \c -boxSizeMult (i.e. independent 00042 of the elements of the distance matrix). 00043 \param randNegEig Picks coordinates at random when a embedding process produces 00044 negative eigenvalues 00045 \param numZeroFail Fail embedding if we find this many or more zero eigenvalues 00046 (within a tolerance) 00047 \param optimizerForceTol set the tolerance on forces in the distgeom optimizer 00048 (this shouldn't normally be altered in client code). 00049 00050 \param basinThresh set the basin threshold for the DGeom force field, 00051 (this shouldn't normally be altered in client code). 00052 00053 \return ID of the conformations added to the molecule, -1 if the emdedding failed 00054 */ 00055 int EmbedMolecule(ROMol &mol, unsigned int maxIterations=30, int seed=-1, 00056 bool clearConfs=true, 00057 bool useRandomCoords=false,double boxSizeMult=2.0, 00058 bool randNegEig=true, 00059 unsigned int numZeroFail=1,double optimizerForceTol=1e-3, 00060 double basinThresh=5.0); 00061 00062 //*! Embed multiple conformations for a molecule 00063 /*! 00064 This is kind of equivalent to calling EmbedMolecule multiple times - just that the bounds 00065 matrix is computed only once from the topology 00066 00067 \param mol Molecule of interest 00068 \param numConfs Number of conformations to be generated 00069 \param maxIterations Max. number of times the embedding will be tried if coordinates are 00070 not obtained successfully 00071 \param seed provides a seed for the random number generator (so that the same 00072 coordinates can be obtained for a molecule on multiple runs). 00073 If negative, the RNG will not be seeded. 00074 \param clearConfs Clear all existing conformations on the molecule 00075 \param useRandomCoords Start the embedding from random coordinates instead of 00076 using eigenvalues of the distance matrix. 00077 \param boxSizeMult Determines the size of the box that is used for 00078 random coordinates. If this is a positive number, the 00079 side length will equal the largest element of the distance 00080 matrix times \c boxSizeMult. If this is a negative number, 00081 the side length will equal \c -boxSizeMult (i.e. independent 00082 of the elements of the distance matrix). 00083 \param randNegEig Picks coordinates at random when a embedding process produces 00084 negative eigenvalues 00085 \param numZeroFail Fail embedding if we find this many or more zero eigenvalues 00086 (within a tolerance) 00087 \param optimizerForceTol set the tolerance on forces in the DGeom optimizer 00088 (this shouldn't normally be altered in client code). 00089 \param basinThresh set the basin threshold for the DGeom force field, 00090 (this shouldn't normally be altered in client code). 00091 00092 \param pruneRmsThresh Retain only the conformations out of 'numConfs' after embedding that are 00093 at least this far apart from each other. RMSD is computed on the heavy atoms. 00094 Prunining is greedy; i.e. the first embedded conformation is retained and from 00095 then on only those that are atleast pruneRmsThresh away from already 00096 retained conformations are kept. The pruning is done after embedding and 00097 bounds violation minimization. No pruning by default. 00098 00099 \return an INT_VECT of conformer ids 00100 00101 */ 00102 INT_VECT EmbedMultipleConfs(ROMol &mol, unsigned int numConfs=10, 00103 unsigned int maxIterations=30, 00104 int seed=-1, bool clearConfs=true, 00105 bool useRandomCoords=false,double boxSizeMult=2.0, 00106 bool randNegEig=true, unsigned int numZeroFail=1, 00107 double optimizerForceTol=1e-3,double basinThresh=5.0, 00108 double pruneRmsThresh=-1.0); 00109 } 00110 } 00111 00112 #endif
1.5.3