Serendipity: Nature's error- correction

CHANCE meetings between researchers from different backgrounds, one of them unwittingly holding the missing piece of the other's jigsaw, can result in significant discoveries. Just such a meeting occurred in a bar at Bradford University, between Dr Simon Shepherd and Professor Terry Baker.

Shepherd, a computer scientist, specialises in techniques for spotting and correcting errors in information. These so-called error-correction techniques are ubiquitous, and are essential for the efficient operation of everything from CDs to the Internet. Even International Standard Book Numbers (ISBNs) have in-built error- correction. The ISBN consists of 10 digits (occasionally one of the digits is an X, which stands for 10), and you can test the error- correction for yourself by looking at any book.

Moving from right to left, take the first ISBN digit, multiply it by one and note the result, then take the second digit, multiply it by two and note the result. Continue in this fashion until the 10th digit, which you multiply by 10. Then add all the results together. The total will be a multiple of 11. Therefore, each time an ISBN is typed into a cash register, the register can do a quick calculation, and if it sees that the total is not a multiple of 11, it knows that the ISBN is wrong and asks for a correction.

In 1994, Shepherd was chatting to Professor Baker, a biologist. Baker was explaining how certain sections of our DNA, genes, carry the information required to build a working human being. However, in between the genes are vast strands of seemingly useless DNA. Baker, along with all other biologists, was perplexed by this so-called junk DNA, because it requires a great deal of energy to make but seems to contribute nothing. One theory was that it contained the leftovers of millions of years of evolutionary dead ends, but experts argued that we were unlikely to carry so much redundant genetic baggage around.

When Shepherd heard about the problem, it was clear to him that junk DNA must perform some kind of error-correction. The useful bits of the DNA are the blueprints for making vital proteins, and Shepherd argued that the junk is there to prevent mistakes occurring during manufacture. This explained why so many genetic experiments had been failing. The intricate activities of genes are highly prone to errors, and genetic engineers had been transplanting genes without the associated error-correcting DNA. As a result, the transplanted genes were not working properly.

However, one mystery remains. The mechanism behind ISBN error-correction is clear but, unfortunately, the mechanism behind DNA error-correction is unknown. Scientists now acknowledge that junk DNA prevents errors, but they have no idea how it works.