A lifetime's supply of chocolate

"I predict that within one-hundred years, computers will be twice as powerful, 10,000 times larger and so expensive that only the five richest kings of Europe will be able to afford one".

The mind boggling rise of computer technology from abstract plaything of Victorian gentry, to 70s technophobe joke, to inevitable robot overloads, is both marvellous and boringly ubiquitous to my generation, brought up with Windows '95 and Netscape.

For people old enough to remember when a Casio digital watch was both a technological wonder and status symbol of conspicuous consumption however, computer technology holds an uneasy place in there life, occupying both the "it's cool to Google my name" and the "it's scary there's a picture of me when I Google my name" parts of their brains.

Two properties which helped stoke this explosion in computer technology were both the plasticity and "scalability" of computer technology. The plasticity of computer tech is the fact that the "bits" of a computer can really be made of anything - in a "normal" computer it's electricity, but it could be anything given enough imagination. Even things which we don't fully understand - such as quarks.



Lego Vs PlayMobil.

 Scalability describes whether a technology is capable of being grown, or whether it's a stand-alone application. A good analogy is Lego. A Lego model can be pulled apart and reduced to it's basic and easily-understood components, i.e. the bricks, and built and scaled up into more complex structures. Compare this to something like PlayMobil - a "standalone" product which can't be reduced and changed or built upon. Computers are like Lego - they have basic parts which can grow in complexity.

One amazing example which illustrates these two properties was unveiled the other week in the journal Science, a scalable computer - made of DNA (Science 113:1196-1201).

This isn't the first DNA computer, but it definitely shows more promise that previous attempts - which were stand-alone constructs. This new computer can solve the square root of any number up to and including 15 (which is 3.87298335 by the way), and it only takes 7 to 10 hours figure this out - I know impressive right?

Being facetious aside, it's impressive that DNA can be cajoled into achieving this mathamatical feat. But further to this is that the design of this DNA computer is eminently scalable as it is essentially based on the same logic used in computers - Boolean logic.



An AND gate.

Boolean logic is the essence behind the earliest computers and electronics - manifest in a technology called logic gates. If you did GCSE physics, you probably came across these as AND and OR gates in circuit diagrams, but there are plently of exotic logic gates like NOR, NAND, XOR and XNOR.

The basic principle is that the gates takes a various "input" signals and then transforms them into something else, an "output signal". An AND gate (pictured) works when the various input signals (labelled as A and B in the image) are both True - or in binary terms, 1. When this is true, the signal coming from the other side of the AND gate (from wire O) will also be True (or 1). An OR gate alternatively has an output of 1 when either A or B (or both) equal 1.

These logic gates can be wired up in large numbers in various sequences to produce some very nasty and complex circuit diagrams which can achieve wonderfully complex computations. The DNA computer was achieved using some of the most basic properties of DNA to produce logic gates.


Compulsary Image of DNA

 
DNA is a double stranded molecule, with the two complementary strands being highly specific for each other. The DNA computer uses this property by creating DNA nanostructures called "seesaw gates".

Essentially seesaw gates consist of pieces of double-stranded DNA which wait around for a signal - a complementary single-stranded piece of DNA - to reach them. Using some DNA-based jiggery-pokery, the incoming single strand (if sufficiently concentrated) displaces one of the strands inside the seesaw gate, and hybridises to the free strand. The displaced single strand of DNA is then a new signal, which can go to another seesaw gate - and so the signal is propagated throughout the circuit.  

Seesaw gates can be tuned into make their properties more interesting. This is achieved by increasing the concentration of the incoming signal DNA required to cause of strand displacement of the gate DNA. By having a few seesaw gates in order, along with a fluorescence-based (and slightly more complex) "reporter gates", we can produce DNA-based NOT and AND gates.

What's great about this method is it's scalibility - just string together the NOT and AND gates and you too can create a DNA computer. What's more is that stringing up about 30 of these seesaw gates can produce a circuit which can produce the square root for 15 numbers - imagine what can be done with 300 or 3000 seesaw gates! Entire fully understood DNA computers could , if robust enough, be used to create synthetic life in an entirely comprehensive and determinate way.