Synthon.h

Go to the documentation of this file.

//
// Copyright (C) David Cosgrove 2024.
//
//   @@ All Rights Reserved @@
//  This file is part of the RDKit.
//  The contents are covered by the terms of the BSD license
//  which is included in the file license.txt, found at the root
//  of the RDKit source tree.
//
 
#ifndef RDKIT_REAGENT_H
#define RDKIT_REAGENT_H
 
#include <string>
 
#include <RDGeneral/export.h>
#include <GraphMol/SynthonSpaceSearch/SynthonShapeInput.h>
#include <GraphMol/SynthonSpaceSearch/SynthonSpaceSearchHelpers.h>
 
namespace RDKit {
class Atom;
class ROMol;
class RWMol;
 
namespace SynthonSpaceSearch {
 
// These are the numbers of bits used in the internal fingerprints.
// The user is not restricted to these numbers for the search.
inline constexpr unsigned int PATT_FP_NUM_BITS = 1024;
 
// This class holds a Synthon that will be part of a SynthonSet.
class RDKIT_SYNTHONSPACESEARCH_EXPORT Synthon {
 public:
  Synthon() = default;
  Synthon(const std::string &smi);
  Synthon(const Synthon &other);
  Synthon(Synthon &&other) = default;
  Synthon &operator=(const Synthon &other);
  Synthon &operator=(Synthon &&other) = default;
 
  const std::string &getSmiles() const { return d_smiles; }
  const std::unique_ptr<ROMol> &getOrigMol() const;
  const std::unique_ptr<ROMol> &getSearchMol() const;
  const std::unique_ptr<ExplicitBitVect> &getPattFP() const;
  const std::unique_ptr<ExplicitBitVect> &getFP() const;
  const std::vector<std::shared_ptr<ROMol>> &getConnRegions() const;
  void setSearchMol(std::unique_ptr<ROMol> mol);
  void setFP(std::unique_ptr<ExplicitBitVect> fp);
  unsigned int getNumDummies() const { return d_numDummies; }
  unsigned int getNumHeavyAtoms() const { return d_numHeavyAtoms; }
  unsigned int getNumChiralAtoms() const { return d_numChiralAtoms; }
  unsigned int getNumChiralAtomsExcDummies() const {
    return d_numChiralAtomsExcDummies;
  }
  double getMolWt() const { return d_molWt; }
  void updateMaxSynthonSetSize(unsigned int newVal);
  unsigned int getMaxSynthonSetSize() const { return d_maxSynthonSetSize; }
  void clearShapes();
  void setShapes(std::unique_ptr<SynthonShapeInput> shapes);
  const std::unique_ptr<SynthonShapeInput> &getShapes() const;
  // Writes to/reads from a binary stream.
  void writeToDBStream(std::ostream &os) const;
  void readFromDBStream(std::istream &is, std::uint32_t version);
 
 private:
  std::string d_smiles;
 
  // Keep 2 copies of the molecule.  The first is as passed in, which
  // will be used for building products.  The second will have its
  // atoms and bonds fiddled with to make them match the product (the
  // aliphatic precursor to aromatic product issue).  The search mol
  // doesn't always work with product building.
  std::unique_ptr<ROMol> dp_origMol{nullptr};
  std::unique_ptr<ROMol> dp_searchMol{nullptr};
  // The pattern fingerprint, used for substructure search screening.
  std::unique_ptr<ExplicitBitVect> dp_pattFP{nullptr};
  // The fingerprint of the dp_searchMol, used in fingerprint similarity
  // searching.  Its type is known by the SynthonSpace that holds the
  // Synthon.
  std::unique_ptr<ExplicitBitVect> dp_FP{nullptr};
  std::unique_ptr<SynthonShapeInput> dp_shapes;
 
  // SMILES strings of any connector regions.  Normally there will only
  // be 1 or 2.  These are derived from the search mol.
  std::vector<std::shared_ptr<ROMol>> d_connRegions;
 
  unsigned int d_numDummies{0};
  unsigned int d_numHeavyAtoms{0};
  // As calculated by details::countChiralAtoms().
  unsigned int d_numChiralAtoms{0};
  // This is the number of chiral atoms excluding those that have 2 or
  // more dummies attached.
  unsigned int d_numChiralAtomsExcDummies{0};
  double d_molWt{0.0};
  // The smallest SynthonSet size this synthon is in, where the SynthonSet
  // size is the number of Synthons in 1 reaction, so 2-4 probably.
  unsigned int d_maxSynthonSetSize{0};
 
  // Once the search molecule has been added, get the connector regions,
  // connector fingerprint etc.
  void finishInitialization();
  // Calculate the number of dummy atoms etc.
  void calcProperties();
};
 
}  // namespace SynthonSpaceSearch
}  // namespace RDKit
 
#endif  // RDKIT_REAGENT_H