Structure registration

Chemical structure registration is an important activity that is necessary for drug discovery. The structures that have been developed by a pharmaceutical company constitute the "crown jewels" of chemical information, and they must be properly and securely archived. The registration process usually involves the process of extracting, cleaning, transforming, and loading the data—sometimes term ECTL. 

Structure registration is the process of entering structural information in a centralized repository, usually a structure database. These repositories serve as a pool for providing structure information that has been created in other departments of a company. Structure databases are set up according to the individual needs of a department or company. They consist of a common representation of a structure in a standardized file format, such as MolFile, SDF, reaction (RXN) (MDL), JCAMP (International Union of Pure and Applied Chemistry), or simplified molecular input line entry specification. Any additional data can be stored with the structure depending on the context typical examples are structure properties, reaction conditions, and literature references. 

When a molecule is presented as more than one topological graph, the individual representations are termed mesomers or, in the context of structure registration, mesomeric duplicates. Mesomers differ only in bond order and placement of formal charges and are used to represent bond delocalisation in a molecule. 

It all goes according to plan we expect to launch ABACUS Version 2 by the end of summer 1987. This would offer the integrated search system, structure display accompanying all appropriate functions on screen and structures optionally included on a range of standard hard copy outputs. Next we shall concentrate on developing Version 3, which will provide direct structure registration by the chemists, stock control... 

The key problems in structure registration have already been mentioned - the treatment of charge mesomerism, tautomerism and stereochemistry and the creation of a unique registry record for each compound. Software solutions to all of these problems have already been found, reported and put into practice in a variety of systems over the last two decades. Given the time constraints under which the Beilstein database creation software had to be completed and put into operation, it was not appropriate to embark on extensive research work to find new solutions to these basic problems instead existing solutions were implemented in a combination which best suited Beilstein s needs. In this sense, the remainder of this paper does not report an5fthing dramatically new, but rather serves to consohdate proven solutions, albeit in a new environment. 

Four chapters on other operational substructure search systems follow. Jean Marcah and co-workers describe the Du Pont global technical information system Steve Welford and Clemens Jochum report on chemical structure registration for Beilstein Online Peter Rusch also covers Beilstein Onhne in a paper about substructure searching on Dialog and the DARC Toolkit which was described by Pascal Huguet is summarised in a chapter written by Bill Town with some suggestions from myself and Pierre Buffet of Questel. 

Chemical data base systems, and especially the DARC Structure Management System (DARC-SMS), have from the beginning offered sophisticated capabilities for chemical structure registration and structure, or substructure, or generic substructure search. 

---