Published online 2 June 2011 | Nature | doi:10.1038/news.2011.343

News

A molecular calculator

DNA square-root solver is biggest molecular computer yet.

dnaPop quiz, hotshot: what's the square root of 13?Science Photo Library/Alamy

DNA holds the code for all living things – and now it can solve square-root problems too. Researchers have built the largest and most sophisticated system yet that performs calculations using DNA.

The design of the latest 'DNA computer' – which uses DNA molecules rather than silicon chips to perform computations — is seen as a significant advance for the field of molecular computing. It demonstrates how biochemical circuits might be built on increasingly large and complex scales, bringing the prospect of long-touted potential applications such as disease detection a step closer.

Consisting of 130 strands of DNA, it is five times more powerful than previous molecular computers. Like a conventional computer, it uses logic gates, which process incoming signals using simple rules. However, these gates are made from carefully designed DNA molecules, not silicon. The input and output signals are also made from DNA, rather than being electrical pulses. The design is published today in Science1.

Although researchers have been making molecular computers since the mid 1990s, this is the first one to be designed using basic rules and principles that will allow bigger and more complex circuits to be constructed in the future. Previous systems were all one-off constructions.

"What's new is we have a systematic way of building biochemical circuits that works, in practice, to build larger circuits," says Erik Winfree, a computer scientist at the California Institute of Technology (Caltech) in Pasadena, who conducted the work with colleague Lulu Qian. "We don't know exactly how this will pay off, but there are groups who are actively engaged in trying to take our circuits and embed them in other chemical environments."

Leonard Adleman of the University of Southern California in Los Angeles, who is credited with inventing molecular computing in 1994, says the result is "another important step" on a journey that is seeing the "gulf between the chemistry of the laboratory and the chemistry of life" being crossed. "I am confident that research in this field will have vast ramifications," he says."But what form these will take remains to be seen."

Roots to the future

The researchers designed their circuit to find the square root of numbers up to 15 and round the answer to the nearest whole number. To calculate the square root, four DNA strands were created that each encoded one of the four digits in the binary version of the number. These input strands were then added to a test tube of salt water containing the system of DNA logic gates, which work together in a chemical cascade to communicate the system's answer using fluorescent colours. A set of four fluorescent colours was used to communicate the two-digit binary number that was the answer. For each digit of the answer, one colour would signal a "1" while a different colour would signal a "0".

The square root was chosen simply to demonstrate the technique, says Winfree. "If you can get chemistry to do something as utterly alien as computing the square root of a four-digit binary number, then you can probably get it to do a lot of other things too," he says.

ADVERTISEMENT

Andrew Ellington, a biochemist at the University of Texas at Austin, is already applying the design. He is looking at how, based on the Caltech team's work, a biochemical circuit could be developed to diagnose malaria – potentially leading to the development of a device that could be deployed quickly and cheaply in the field. It could calculate its answers using chemicals found in blood. "We can look at the path Qian and Winfree used and design off that," he says.

Ehud Shapiro's team at the Weizmann Institute of Science in Rehovot, Israel, who developed a molecular computer in 2004 that was theoretically able to diagnose cancer, also praised the work, describing the molecular computing system as "astonishingly complex" compared with existing systems.

But others noted the system's limitations, such as its speed of execution (it can take up to 10 hours to compute a square root) and the difficulty of achieving success in the complex environment of living organisms. "The biggest challenge will be to get this type of construction to work inside living cells," said Martyn Amos, an expert in DNA computing based at Manchester Metropolitan University, UK. 

  • References

    1. Qian, L. & Winfree, E. Science 332, 1196-1201 (2011).
Commenting is now closed.