Information Systems Upjohn -. COUSIN

Earger chemical and pharmaceutical firms have, over the years, developed in-house systems with capabilities that are specific to the chemist s needs. Today, the costs of developing from scratch and maintaining an in-house system are prohibitive, especially because commercial chemical information systems are highly efficient and customizable. Personal chemical information software is still being developed and reported in the literature. Examples include a relational database patterned after the Upjohn Cousin system (108), and CheD, which is a SQL-based system with a Web client (109). 

A number of commercial systems that were available at the time were evaluated. Several failed to meet all criteria, due to a lack of U.S. installations and/or U.S.-based system support groups. Those that failed were DARC/Questel, Telesystemes, Paris, France the NIH-EPA Chemical Information System CHEMPIX, developed by Moreaux of Roussel-Uclaf and marketed by Chemical Information Management Inc. and SYNLIB, marketed by Smith Kline Beckman. The Upjohn COUSIN system, developed by Dr J Howe, was not commercially available. Chemical Abstracts Service (CAS) private file support was considered too expensive. Security with the CAS file was also a concern of MSDRL since it could not be brought in-house. 

This paper proposes extensions to the relational database model that allow chemical structure and other complex data types to be included in a relational database. It is argued that this approach provides benefits over the common practice of storing chemical structures in a chemical database system and associated research data in a relational or other general database system. The design, implementation, and usage patterns of an extended relational system are discussed in the context of the Upjohn Cousin compound information system. Emerging extensibiUty features that support the proposed approach within commercially available database systems are reviewed. 

The evolution of molecular graphics (7) is described in an earlier ACS Symposium Series book (77) which acts as an interesting precursor to this present volume. Chemical reaction systems such as LHASA (Logic and Heuristics Applied to Synthetic Analysis) 18) and SECS (Simulation and Evaluation of Chemical Synthesis) (79) had long used graphics but it was some time before the first in-house, proprietary system appeared, attracting much interest in the chemical and pharmaceutical industries. This was Upjohn s Compound Information System, COUSIN 20-21). 

In-house chemical information management systems began to emerge at some of the larger chemical and pharmaceutical firms. These included CONTRAST and SOCRATES at Pfizer, SYNLIB at SmithKline, COUSIN at Upjohn, MSDRL/CSIS at Merck, and CROSSBOW at ICI (27). The Chemical Abstracts database was made available online in 1967 (28). In 1980 this became CAS ONLINE. A compre-... 

In the maximum dissimilarity (MD) selection method used at Pharmacia Upjohn [6], the first compound is selected at random and subsequent compounds are then chosen iteratively such that the distance to the nearest of the compounds already chosen is a maximum. The compounds are represented by COUSIN fragments [13]. Polinsky et al. [42] use a similar algorithm in the LiBrain system. In this case, the molecules are represented by a feature vector that contains information about the following affinity types aliphatic hydrophobic, aromatic hydrophobic, basic, acidic, hydrogen bond donor, hydrogen bond acceptor, and polarizable heteroatom. 

By 1984, the use of chemical database systems was well understood and wide-spread in the chemical/pharmaceutical industry. Their use provided researchers with rapid access to company-wide, and in some cases industry-wide, chemical information. A few companies had developed proprietary systems for this purpose such as Upjohn s COUSIN and ICFs CROSSBOW. However, most companies were using commercial chemical database systems such as Molecular Design s (MDL s) MACCS. ... 

---