effectiveness. Did we address the right problem? Did the information packet cause any different action on the part of the recipient? Was the information even used in subsequent action steps by the recipient? What was the ultimate value of the decision to which the information contributed? Was its contribution to the decision substantive or supportive? Was it filed away for possible future use or tossed into the round file?We first have to do the right thing and then learn to do it well. Packets of information (large or small), even delivered on time in neat array, with great eye appeal, have no value unless they are used to solve problems or support productive decisions. Hence, we do need an integrated approach with our users. We need candid feedback regarding the usefulness (value if you will) of the data/information/knowledge transferred.
SUMMARYAn overview of the environment (from my perspective) related to information resource management in pharmaceutical R&D has been presented. Some notions of organizational preference (functional), employee selection (chemist turned information scientist), automation (user friendly, cost effective), and the value of project teams (information transfer) have been noted. Difficulties associated with keeping our innovative tools sharp were observed. Finally, we noted that our bottom line-productivity-should first consider what is useful (effectiveness) and then learn how to do it wellefficiency. Success in the management of information resources depends on the proactive delivery of information packets which find their way into problem solving and decision support for scientists or line managers.A representation scheme for generic chemical structures (Markush formulas) is presented. This method is an extension of a BCT (block-cutpoint tree) representation for specific chemical structures, and is called an EBCT (extended BCT) representation of generic chemical structures. A general Markush formula is dissolved into several simple Markush formulas by expanding conditions and nested substitutions. Variable parts of simple Markush formulas are represented in terms of a simple type of substituent group called a generic unit. An inverted file for homologous series and a fragment screening facilitate (sub)structure searches.
INTRODUCTIONOne of the important problems in dealing with chemical structure information by computers is how to implement the representation of chemical structures in computer storage. So far, many representation methods have been proposed and implemented for specific chemical structures.' Those methods are categorized to two major approaches: the linear representation and the topological representation. One major reason why so many representation methods have been and will be devised is that we do not have such a complete representation method yet to support various types of substructure searches.