My elementary-school kids have no idea what a typewriter is, and they like to hear how I didn’t even see a computer mouse until I was in college. The story is especially amusing to my first grader, who has to explain to me how the Smart Board in her classroom works. There’s no doubt we are preparing the next generation for jobs that haven’t been invented yet. So thinking about how the next generation of drug therapies may have their start, not in research laboratories but rather within computer databases is exciting.

The University of Michigan is hoping to receive a $5 million grant from the National Institute of General Medical Sciences to help fund a program that will boost its Web database of molecular information. The information will be accessible to scientists and can be used to develop computer-aided design programs designed to predict potential drug candidates.

Computational tools have been part of research and development for some time, but they are only as good as the database on which they are founded. So far, researchers have needed to better understand the structures of and binding strengths between proteins and drug targets.

Building the new discovery resource will be a formidable task, requiring information collected from literature, existing databases of protein-ligand complexes, and newly generated data. Some pharmaceutical companies also have agreed to contribute unpublished data, and researchers are holding events to encourage industry collaboration to fill in any gaps in the database.

Of course, in time the database will mostly likely be as obsolete as the college typewriter, but by then we hope today’s diseases also will be.