After my last post on alternative biochemistries, I thought I’d link to something at least tangentially related, and talk a little about something that’s almost the reverse idea: trying to use the existing DNA information system in our cells to replace silicon-based electronics in computing.
I recently read Christ Moriarty’s Spin series, which I enjoyed a lot. I’d recommend them to anyone interested in well-written SF with a terrific protagonist, well-drawn worlds and characters, and interesting plots. The science was a little…speculative at times (probably more so to a physicist) but there was only one bit that gave me a real facepalm moment, in the last book (no, this isn’t really a spoiler), which describes running an AI (as in thinking Artificial Intelligence) in a human body by using their DNA as some sort of operating system (I wasn’t clear how). Um, no. The main No being that the human body in question is using it, actually (and not, strictly speaking, as an “operating system”). If you made changes to the sequence then you’ll mutate it and that body will cease to function, so you can use the DNA as neither hardware nor software. (For a good takedown of the “DNA as a computer analogy, which I can’t help but think muddles things more than it illuminates, read this blog post here).
Apart from all that metaphorical talk of DNA as a computer, etc., the idea partly originates because DNA potentially offers massive parallel computing power. However, we’re talking chemical reactions here, which are going to be very slow, compared to electrons hopping about silicon chips. There have in fact been some experiments in making DNA-based computers, including a sort of biological transistor, with the aim of exploiting that processing potential to solve complex problems, but these are constructed entirely on their own, not as something somehow embedded within a living organism.
Where DNA is currently showing more promise in the field of computer science is in fact as information storage. Only a couple of weeks ago, I was telling my students how every cell in our body contains about two metres of DNA squeezed into something that has a diameter of only a few tens of microns across (and that needs room for other celly things!). DNA can package in three dimensions, and very efficiently. Smarter minds than mine have noticed this potential, and have been working on creating methods to use DNA to store data, with some success; Microsoft is investing in research in this area, and it’s being noticed in the technology sector news too. There’s a nice rundown of the theory behind a recent success here.
A recent Nature News piece had a nice graphic illustrating the potential superiority of DNA in terms of storage capacity compared to current methods:
This isn’t just for the sake of academic interest: the limits to data storage are becoming an increasing problem, and the world generates more data every year.
The obvious caveat is that this DNA-based storage could only be used really for archiving: you’re not going to get instant access data retrieval when you need to synthesise new chemicals to get your data each time. Still, it’s a rather fascinating idea, and I’d be interested to see how this area of research develops.