Chemical substructure searching on CD-ROMs

Abstract

Substructure searching is one of the most valuable tools for retrieving a particular class of compounds from chemical databases. Other search techniques using molecular formula, atom counts, ring indices, name fragments or any physical or chemical properties are not able to precisely retrieve compounds with a given structural property. Molecular formula and ring descriptions reflect only a very special feature of the structure. Chemical names, if correct, describe a structure unambiguously. However, due to the fuzzy nomenclature rules it is almost impossible to expect a user to define a query with a chemical name that is identical to the name stored in the database, not to mention the language difficulties. The same is true when using chemical name fragments, in fact there is not a one-to-one relationship between structural fragments and name fragments. This problem is even more severe when searching databases where each compound has not got a name, or has only one of a large number of frequently used synonyms.