Digital design is set to revolutionise biochemical analysis, according to US scientists.

James La Clair and Michael Burkart from the University of California, San Diego, have looked at ways of computerising high-throughput assays to make them more accurate and useful. High-throughput assays are used widely in industry and medicine for rapid analysis to screen new drug candidates or to diagnose disease quickly, for example.

La Clair and Burkart studied recent assay designs based on CDs. These assays look for a target analyte by collecting light reflected from or emitted by molecules attached to the surface of the CD. When an analyte binds to one of the surface molecules, there is a change in the way that molecule emits or reflects light. Detecting this change reveals the presence of the analyte.

Detecting and correcting errors in the signals from these types of assays is very hard. This means a large number of repeat assays must be done to be sure the results are correct. This is often done by putting many copies of the assay onto the CD.

La Clair and Burkart advocate a different way of designing assays. Their method starts with a known pattern of surface molecules that creates a repeating background pattern in the signal. Changes in the signal indicate that an analyte is present.

The key advantage of this technique is that the regular pattern of surface molecules forms a digital signal that can be fed to a computer, making the processing and interpretation of the signal much faster and easier. Because the background signal can be used to tell what position on the CD the signal comes from, the assay results can be read out more accurately and fewer repeat copies of the assay are needed, saving space on the CD. 

La Clair and Burkart hope their method will lead to a new generation of molecular-based digital machines which could provide unique advantages for biomolecular science.

Clare Boothby