University of Rochester Science News


May 21, 1997

A pair of scientists at the University of Rochester has built some of the first DNA computer "hardware" ever: logic gates made of DNA.

Most surprising about Animesh Ray and Mitsu Ogihara's recent work is that they made the DNA logic gates using only the most commonplace biological laboratory techniques, such as DNA ligation and gel electrophoresis.

"There is absolutely nothing fancy in what we have done," Ray says. "The techniques we've used are the same ones that thousands of biologists use every day in their labs." Ray, an assistant professor of biology, and Ogihara, an assistant professor of computer science, took fellow researchers by surprise with their announcement at the recent First International Conference on Computational Molecular Biology in Santa Fe, N.M.

Ray and Ogihara's logic gates rely not on electrical signals to perform logical operations -- as do logic gates in today's computers -- but rather on DNA codes. Ogihara says that it could be several years before the DNA logic gates are actually incorporated into a working DNA computer. In the meantime, he and Ray will explore the conditions that allow for the most accurate and efficient DNA computation.

The building blocks of all today's computers, logic gates are tiny structures that convert the endless series of binary data coursing through every computer into a series of signals that a computer uses to perform its operations. Today's logic gates process electronic signals from transistors made of materials like silicon, converting two input signals into one output signal in a way that allows a computer to perform complex operations. Up to now, the only logic gates used for computing have been electronic structures that detect signals coming from transistors.

Ray and Ogihara's DNA logic gates open up whole new possibilities, since they mark the first step toward building a DNA computer that would perform calculations in the same way as electronic computers. These gates -- actually tiny DNA processing centers -- detect specific fragments of the genetic blueprint as input, then splice together the fragments to form a single output. For instance, a genetic And gate links two DNA inputs by chemically binding them so they're locked in an end-to-end orientation -- much as two Legos might be fastened end-to-end by a third Lego stuck on top. An enzyme called DNA ligase seals the gap between the ends of the two input strands, yielding a single new strand.

Using regular gel electrophoresis, the length of this new strand can be precisely measured, providing the DNA computer's "answer" or output to the two input strands. For example, with the DNA And gate, a final DNA sequence that's as long as the input strands linked together indicates that an And operation has occurred. A similar system of two short DNA strands work together in Ray and Ogihara's Or gate, whose answers are likewise read in the length of the resulting DNA strands.

Most researchers have focused on possibly building DNA computers to tackle killer problems that traditional computers cannot solve, such as the famous "traveling salesman" problem. The Rochester team is one of the first to seriously consider whether DNA computers might be used for problems now routinely done by electronic computers, and to emulate the way electronic computers "think."

Ogihara recently showed mathematically that a computer consisting of a series of DNA-filled test tubes can work more efficiently than a digital computer in analyzing the information cascading in from a tangled web of logic gates. This includes the type of calculations now done every day, as well as more complex arrangements. A DNA computer would need just a few hours to analyze a flood of information that would take today's conventional computers hundreds of years to solve.

Ogihara and Ray believe that when coupled with new DNA microchips being developed by researchers at a biotechnology company called Affymax, the gates will usher in a wave of breakthroughs in DNA computing. Ray says such chips could make DNA computation even faster by speeding the tally of the DNA strands that serve as answers to computations. Rather than running the DNA through a slow gel electrophoresis, researchers could add labeled strands to a DNA chip, which consists of hundreds of squares containing different known strands of DNA. The added DNA would quickly bind to the strands in the square containing its complementary DNA sequence, and scientists could use the labels to detect the DNA answers.

Ray and Ogihara are among a growing group of scientists who believe that DNA could serve as a very compact, efficient, and accurate form of memory in computers -- just as it does in the cells of the human body. The potential benefits of a DNA computer are astounding: One pound of DNA has the capacity to store more information than all the electronic computers ever built, and the computing power of a teardrop-sized DNA computer using the new DNA logic gates could dwarf that of the world's most powerful supercomputer, which is the size of a house. Ogihara believes DNA will prove useful on problems that now require vast networks of computers, such as forecasting the weather, designing airplanes, or cracking complex security codes.

CONTACT: Steve Bradt, (716) 273-4726.
Latest News at the University of Rochester