BIOCOMPUTERS will never replace electronic machines, despite last week鈥檚 news that a DNA computer has solved the biggest problem it has ever attempted. But they could have a valuable role inside the human body.
Ravinderjit Braich and Leonard Adleman, the scientist who created the first DNA computer in 1994, used biological molecules to solve a complex logic problem called 3-SAT. The researchers, both at the University of Southern California in Los Angeles, synthesised more than a million strands of DNA, each 300 bases long, to encode every one of the possible solutions to the problem. In a series of logical steps, they eliminated successive batches of strands from the original pool of possibilities until the final result emerged.
The calculation, which will be reported in a future issue of Science, took several days to complete. A conventional computer would have taken seconds. 鈥淲e concluded that our methods would not allow us to outperform electronic computers on classical computations,鈥� says Adleman.
Advertisement
One of the major problems they encountered was the 鈥渕isbehaviour鈥� of DNA molecules during the process. 鈥淥ver the past few centuries electrical engineers have learned to control electrons with exquisite precision,鈥� Adleman says. 鈥淯nfortunately chemists and biologists have not learned to control biomolecules with anywhere near this kind of precision.鈥� He says he doesn鈥檛 believe a breakthrough is on the horizon. 鈥淏ut I鈥檓 anxious to be proven wrong.鈥�
鈥淎dleman鈥檚 approach will never compete with electronic computers,鈥� agrees Yaakov Benenson of the Weizmann Institute of Science in Israel. But while direct competition with conventional machines may be futile, Benenson and his supervisor Ehud Shapiro believe that DNA computers will have the edge鈥攊nside the body.
Benenson and Shapiro recently built a computer in which DNA strands are manipulated by enzymes. 鈥淲e鈥檙e trying to build devices that have biomolecules as inputs and outputs,鈥� says Shapiro. Eventually they hope to inject tiny DNA computers into cells. These could detect and identify genetic abnormalities, such as changes that lead to cancer, and produce molecules to block the rogue genes.