DNA logic gate components……Credit: Royal Publishing Society
Let’s come at computers from a different angle for a moment. An alien species lands on earth. Their spaceship doesn’t look like a spaceship. It looks like a very large blob, of sorts. It’s a blob because the whole thing is organic, not a scrap of metal on it or in it. The aliens are, of course, also organic. Their entire technology uses only organic compounds, as we call them (praise be to carbon). That means their on-board computers are also organic. No silicon chips. In fact, they don’t even use graphene, which is pure carbon but not organic. Instead they use computers based on DNA and related substances, which with minor variations is almost the same as our DNA. They have used DNA because on their long voyages through space only DNA is self-perpetuating, self-repairing. Their computers get old, but a new crop is always in preparation. Their organic computers are infinitely recyclable and require only the minimum of manufacturing capability.
Okay, that’s science fiction. Computers based on DNA are not science fiction, quite. Biologists and computer scientists have been attracted to the notion of using the combinatorial power of DNA to perform computer-like calculations for decades. It’s fair to say though that moving from what seems to be a logical use of DNA to the actual biological material has not been easy. No human made ‘biological computer’ exists; the problems are far too complicated. These days what most scientists choose to do is work on something comparatively simple – logic gates, the basic computer component, built from DNA.
Lulu Qian and Erik Winfree with their team at the California Institute of Technology (Caltech, Pasadena, USA) built the first such logic gate, or DNA circuit, in 2006. They then constructed a 5-layer 12-DNA molecule circuit from these gates – and the processing speed fell off a cliff, orders of magnitude slower. Back to the computer modeler, as they say. The result is a new design, published in the June 3, 2011 issue of Science and available in a version without paywall at Royal Society Publishing 6 June 2011, [A simple DNA gate motif for synthesizing large-scale circuits] One of the circuits built with the new approach used 74 different DNA molecules, the largest such circuit to-date, which can calculate the square root of numbers up to 15. Yes, obviously this is not even a crude silicon calculator, but that comparison misses the point entirely.
First, I’ll try to get to the nub of how this DNA gate (or circuit) works: Qian and Winfree have constructed several kinds of circuits, but perhaps the type most easily understood are logic gates NOT, AND, OR. (Sorry about the computer jargon, but such logic gates are fundamental to computer operations.) Let’s start with the idea of a double stranded helix of DNA, not the whole long chromosome but a short piece of one. It contains genetic code, the base pairs of adenine-guanine and cytosine-thymine (A=G, C=T) in a sequence that the researchers can determine and create – that is, program. A piece of double stranded DNA like this is built so that it has an ‘open section,’ an area where other pieces of DNA (with appropriate bases) can attach. This is the basic DNA gate, which typically is immersed in a salt water solution.
To make it work, the researchers create a marked ‘output’ molecule that will be preloaded and attached to the gates. Then ‘input’ molecules are introduced to the solution. One by one, the input molecules, which may have been created with a greater affinity to bond with the DNA gate, displace the output molecules. Eventually most of the input molecules are bonded to the gate and a count can be taken of the output molecules.
By combining two such DNA gates and programming their level of output, when both gates are not bonding (off) with input molecules, their output molecule count is low (NOT); if one is on and the other is off, the output is higher (OR); if both gates are fully bonding, their output is high (AND).
This is oversimplified, but the point is to visualize that this system uses molecules as input (programmed and counted) bonding to a DNA gate, which in turn releases output molecules (also programmed and counted) to determine a result. Schematically this is a pretty simple system, although chemically it is sophisticated. The researchers consider this a great strength, as the basic approach can be used to create composite circuit constructions without radically increasing the level of complexity. In the parlance of technology, this approach can scale (grow without collapsing).
Using this basic approach, the researchers have been able to create several different types of gates. They’ve gotten to the point where if you can tell them what logic or computational effect you need, they can ‘compile’ the correct DNA sequence for the gates.
At this point, researchers can demonstrate that this approach works; but it is slow, requiring ten hours to do the calculation of a square root. That’s why I say it’s not helpful to compare this to silicon-based computing. What’s important here is that as an organic system, this kind of ‘computer’ can be seamlessly interfaced with other organic systems. With the right packaging, a DNA computer of this type could operate within living cells, or certainly within living tissue and use organic molecules from cells as the inputs or outputs. However, that is a long way in the future, perhaps a decade or more. In the meantime, researchers will be looking for ways to speed up the process; make it more readily programmable, and of course increase the capacity and complexity of the computation.
It’s too early to say if the approach taken by Qian and Winfree will ultimately succeed in becoming a practical DNA computer; but it’s not too early to say that as a ‘proof of concept’ it points the way toward another mode of computing with enormous (good/scary) potential.
“Like Moore’s Law for silicon electronics, which says that computers are growing exponentially smaller and more powerful every year, molecular systems developed with DNA nanotechnology have been doubling in size roughly every three years,” Winfree says. Qian adds, “The dream is that synthetic biochemical circuits will one day achieve complexities comparable to life itself.”