Milestone (1985)

Mainframe computers like those shown here once stored structural, inventory, and shipping data. Courtesy of NCI. 1980.

Drug Information System (DIS)

From 1955 to 1985, structures for compounds being submitted to DTP for evaluation were hand-drawn, at first on Chemistry Cards and later transferred to a computerized database by the Chemical Abstracts Service (CAS). In 1985, a contract effort to develop an electronic database eliminated the need to hand-draw structures. DTP now had an in-house Drug Information System (DIS) that allowed the transfer of input activities from CAS to the acquisition contractor and gave the staff the capacity to perform interactive searches of structural, inventory, and shipping information. It is fair to state that the level of enthusiasm for learning how to use the DIS varied among the staff. The software resided initially on an NIH DEC system 10 mainframe computer, which, when in full use, was painfully slow. Basically, the DIS served as a storage and retrieval mechanism for both current and historical data. It had two major subsystems termed "Pre-Registry" and "Drug." The Pre-Registry subsystem handled compounds suggested for acquisition, while the Drug subsystem stored data on drugs accepted and received for anticancer evaluation

The DIS, when developed in the early 1980s, was built using FORTRAN, which was the best programming language available at the time for the application. By the mid-1990s, relational databases were becoming more user- and developer-friendly, and an effort to create an updated system to handle the needs of DTP was begun using an ORACLE-based platform. The effort was extensive in that the software would be required to handle not only inventory, shipping, and supplier information, but chemical structures and biological information as well. The system had to be appropriate for storage, retrieval, and searching functions for all databases that had grown over the almost 45 years of the program, which complicated the task. There was also a large learning curve in moving from the FORTRAN-based system to the new system. Thanks to the efforts of the Information Technology Branch and some talented and patient contract staff, the new system, called OMNIS, became the working system in 2005.