Substructures Searching Distance

Set reduction involves the successive elimination of candidate structure atoms from sets corresponding to each pattern atom on the basis of an analysis of neighbourhood and connectivity information. The technique has been widely used as a component of 2-D substructure searching systems and we have developed a modification of the technique which can be used for geometric searching. The first stage of the algorithm involves the creation of a distance table. The NQ pattern atoms are labelled from 1 to NQ and for each of the NQ(NQ-l)/2 distinct interatomic distances in the pattern (or less if not all of the query distances are specified for the search), a list of pairs of atoms from the structure is produced. The distance between the atoms in these pairs is equal to that between the pattern atoms (to within any specified tolerances), and the atom type of the first atom corresponds with the type of the first query atom (and similarly for the second atom). 

Substructure searching, whether in 2D or in 3D, provides an invaluable tool for accessing databases of chemical structures. It does, however, have several limitations that are inherent in the retrieval criterion that is being used, which is that a database record must contain the entire query substructure in precisely the form that it has been specified by the user (subjea to any variations that have been specified, e.g., a distance tolerance or a range of acceptable substituent types for a specific position on a ring system). 

Firstly, and most importantly, a substructure search requires that the user who is posing the query must already have acquired a well-deffned view of what sorts of structure are expected to be retrieved from the database. For example, a 3D substructure search requires sufficient information about the geometric requirements for activity to be able to specify distance and/or angular constraints to characterize those molecules, and just those molecules, that can fit some putative receptor site. This implies that some form of pharmacophore map has been created, using techniques such as those described by Bures et al., which in turn implies that it has already been possible to identify sufficient active substances to generate the map. Substructure search is much less appropriate at the start of an investigation when perhaps only one or two active structures have been identified and when it is not at all clear which particular feature(s) within them are responsible for the observed activity. 

---