The computer that haunts Martyn Amos’s dreams literally comes to life when he switches it on. With the potential of awesome computing power it bears little resemblance to a conventional machine. Where you might expect silicon chips to lurk is a noodle soup of cells, chemicals and DNA strands. And to keep it going, Amos has to ensure it is supplied with nutrients, not electricity. For the computer of the future will not be a mass of inert circuitry. It will be alive.
It’s a vision that fires Amos’s imagination. A lively young academic based at Manchester Metropolitan University, he has become a leading figure among scientists intent on creating a new sort of biocomputer with the potential to solve a range of problems that their conventional, inert counterparts struggle with.
But these scientists do not want to stop with living computers. They wish to go on to reprogram life itself, to create new microscopic life forms that tackle disease, clean up pollution and myriad other applications.
They call this synthetic biology and eagerly anticipate it becoming one of the frontiers of 21st century science. More fundamental than genetic modification, the project is barely a decade old and has received little public recognition.
At the heart of this effort sits DNA, the coiled molecule that stores the information necessary for the life of every living organism, from plankton to pandas to palm trees. Through billions of years of evolution, this remarkable chemical has come up with answers to every challenge our planet has thrown at it. The prospect of harnessing the computing power that has helped DNA to thrive has long inspired scientists.
But it is no longer just a dream. The first man-made DNA computers have already proved themselves in action. And while limited in the questions they can tackle so far, the machines have sparked enormous excitement in the scientific world.
Amos decided to build one at Warwick University in 1994 as he embarked on the world’s first PhD in biocomputing. At that time only one such experiment had been carried out. The American scientist Len Adleman — the man who invented the main internet encryption system — had just created a stir by announcing he had used molecules of DNA to make calculations.
At Warwick, Amos supervised the creation of a computer that comprised a pot of DNA and enzymes. Its success added to the excitement, but the limitations, too, were apparent. While DNA is really effective at solving problems in which there are an enormous number of possible answers, practical constraints rule out any takeover from silicon.
DNA computing works by exploiting the way the molecule stores information and interacts with chemicals called enzymes inside living cells. Separate strands of DNA can be tailored to produce different sequences of genetic material representing every possible answer to the problem you want to solve. The piece that represents the right answer then has to be fished out.
Amos does this through a process of elimination, using enzymes designed to rip apart swathes of wrong answers until the right one remains. It then signals the result through another chemical reaction such as one that that creates a flash of light.
It is this ability to make a vast number of calculations simultaneously — each one asking whether a particular strand of DNA is the wrong answer to the question — that is the secret to DNA computing’s prowess.
Each step in biocomputing is relatively simple — any graduate student could do it, Amos admits. The knack is learning how to phrase questions in the language of DNA and its enzymes. In the past decade, huge advances have been made, building on Adleman and Amos’s early DNA computers.
The sophistication of the machines is growing fast — one is now unbeatable at noughts and crosses (though it insists on having the first go). By hijacking the way nature “computes”, Amos says, “researchers in the field of biocomputing are looking to force a fundamental shift in our understanding of computation”.
But that is not enough. The scientists want to blur the boundaries separating biology, chemistry, engineering and computing still further. They want to harness DNA not only in the Petri dish but in its natural environment.
In the living cell and given the right conditions, DNA is capable of performing every trick that evolution has taught it over 3½ billion years. If you could attach that vast armoury of ability to the power of your DNA computer, you have the enormous potential of the cellular computer.
The flood of genetic information coming from a host of genome sequencing projects that not only include humans but many other organisms from onions to orang-utans is like a catalogue of genetic components for synthetic biologists. In time they hope to pick what they need from such databases to build molecular machines for specific tasks. For instance, a harmless bacterium could be modified into a microbot, programmed to sniff out the chemical traces of a newly formed cluster of cancer cells and emit a molecular signal to wipe out the diseased tissue.
Such cellular machines are a long way off but more limited, modified DNA computers are showing the way ahead. One has recently been built to signal the presence of prostate cancer in a test tube while another is being constructed to test drinking water for arsenic contamination, a problem that affects millions of people worldwide.
The ultimate goal is to build cellular machines from scratch. But unlike the world of silicon, biology is a messy, imprecise field. Genes interact in complicated ways and tinkering with them piecemeal can have unexpected effects. But while such doubts have plagued the work of genetic engineers, the synthetic biologists want to embrace the messiness, assembling their living machines piece by piece like some flatpack furniture kit.
“This is just the beginning,” Amos says. “The guys who built the first transistor could never have imagined the massive impact their invention would have on the world. Synthetic biology has the same potential.”
Dr Martyn Amos will be discussing his new book, Genesis Machines: The New Science of Biocomputing, at London’s ICA this Tuesday.
