
The chemical libraries of large pharmaceutical companies typically comprise a million or so compounds, all of them typically stored in impressively large, vault-like structures. But what a different picture appears here: four small vials resting in the palm of a hand. Yet they contain some ten billion compounds—yes, that’s “b” for billion.
The vials are to be found within Screening and Compound Profiling, a part of GlaxoSmithKline that starts the R&D process by identifying compounds which bind to drug targets. The “hits” emerging from this screening are then chemically optimized to have more drug-like potential, more stability, for example, or still more affinity for the target.
The pioneering work, done within a small group at GSK-Boston, has to do with something called Encoded Library Technology (ELT). It creates multiple, ultra-large libraries of drug-like structures through chemical synthesis. Crucially, each structure is encoded, or tagged, with a unique sequence of DNA during the synthesis process. The libraries are then pooled and rapidly screened to determine which structures or families of structures interact with selected drug targets. Since each structure is encoded, it can be easily identified if it is found to interact with the target.
Because ELT is currently suitable only for screening certain kinds of targets, it serves to complement, not to replace, GSK's automated high-throughput screening. ELT is nevertheless increasing the number and diversity of compounds brought to drug discovery. Given the long odds against any compound, taking more “shots on goal” in this way promises to yield more and better leads for new medicines.