Signal transduction is in some sense like the great arduous physical feats of exploration and mapping: What at first sight appears to be straight passage across the Rockies or the Alps turns out to be an endless expedition that in some instances finds itself back where it began. The cell's signaling pathways turn out to be more reminiscent of the massively redundant, overlapping neural networks of the brain than neat Manhattan traffic grids of interacting streets and vehicles. Understanding the interactions between these pathways is intrinsic to understanding how all cellular activities are regulated.

Biotechnology took up signal transduction as a drug discovery platform 15 years ago when Stanley Crooke and George Poste organized the first pharmaceutical symposium devoted to cell signaling in the context of the new recombinant genetic technologies at Smith Kline and Beecham in 1984. It was a marvelous and seductive first sighting: All physiological functions and malfunctions depend on signal transduction. Figure out what happens at the cell surface when a cell receives a signal, and then map out how it transmits that information to the various cellular machines controlling gene expression, and you will succeed in identifying any number of drug targets and points of therapeutic intervention. Too much or too little signaling should also reveal therapeutic endpoints. But then the signal transduction expedition turned into the new particle physics expedition—a blizzard of entities with whimsical names and acronyms blew up, the hoped–for linear pathways turned out not to be linear, and many different potential targets turned out to use very similar or overlapping pathways, making accurate target selection difficult, and the development of a "big picture" even more so.

Despite the difficulties it presents, and despite the insurgence of upstart genomics, signal transduction is still in many quarters the royal road to the New World of Therapeutics. Every major pharmaceutical company has a signal transduction drug discovery program, and more than 30 smaller biotechnology companies are pursuing discovery research via this route (see Nat. Biotechnol. 16:1082, November 1998). The research coming out of this area is copious and rich (see, for example, Redfern et al., p. 165), and several promising signal–transduction–based compounds are in clinical trials. The fact that more than 800 delegates plan to attend the Nature Biotechnology Miami Winter Symposium on signal transduction and therapeutic strategies this month is testimony to the interest in all things transductive.

But what's clear from all of signal transduction's hard work is that a single field of research will not likely turn out to be the "ideal" drug discovery platform on its own. The application of powerful high–throughput genomic and proteomic approaches to signal transduction will yield more satisfying results (see Nat. Biotechnol. 16: 1329, December 1998). Indeed, the integration of biochemical and molecular approaches, genomics, proteomics, and bioinformatics will be needed to fully illuminate the complex interactions in and between these pathways and to analyze regulation at many levels simultaneously—so there is plenty of opportunity for collaboration on all sides. It also seems clear that sophisticated computer modeling of these networks will be needed to successfully develop the "big" picture. Advances in computation should allow us to see how the system is put together in the broadest sense—where else and how quickly might the intrepid cartographers of yesterday have gone with a computer and a satellite to extend their already formidable senses!