Patents are a valuable and often under-utilised information resource. They often have substantial chemical content, much of which may never be published elsewhere. Because patents are legal documents, the authors usually try to make the description of the chemistry as broad as possible, using generic structures which can represent an almost infinite number of chemical compounds. For this reason, patents present problems (challenges!) for information systems unlike those found in other areas. During the 1980's considerable time and effort was put into solving the 'Markush Problem'. From this work, two commercial services, MARPAT and Markush DARC emerged. These Markush search systems have remained more or less unchanged, and their use has been mostly limited to searches directly concerned with intellectual property. In contrast, developments in storage and retrieval of specific chemical compounds have continued, including recent enhancements prompted by the advent of combinatorial chemistry to handle certain types of generic structures.
We would like to discuss the business problems raised by generic structures, ways in which recent developments may help to solve the problem, possible overlap with combinatorial chemistry and how we can make this information more accessible and useful.