Blog: Dan E. Linstedt

« Architecture, Standards, and Business | Main | ELT and ETL - candid view of pros and cons. »

DNA Robots and Computing Technology

I usually don't use a title that someone else has used, but I feel that this is a VERY important breakthrough. See this site for the story I'm blogging on: http://www.jefallbright.net/node/2616

In this blog I will go on an exploratory journey into what it is like, what it would be like to establish computing power on the DNA level. This again, is conjecture - pure speculation, so it's ok to let your mind wander a bit.

In one of my recent blogs, I predicted (not the first to say this) that DNA computing appears to be the strongest and most rapidly advancing field (in terms of nanotechnology applied to computational ability). There are other parts of nanotech that are advancing rapidly as well (in other areas, like bio-informatics).

This is quite astounding. Man made atomic level substances built to do something specific, walking around in a liquid solution - attaching to other molecular tracks and actually "walking" across the strands. All in parallel, all atomic layer control. Fascinating.

Now let's step off the path for a minute and see what this might lead to in terms of computational power. Let's ask a few questions, and hope that someone in this field (familiar with this technology) will comment for us.

The first question that comes to my mind is: why was the wheel re-invented? In other words, there are enzymes that travel down a single DNA strand and unzip it, there are other enzymes that travel down the same DNA strand and zip it back together, there are additional enzymes that replicate DNA strands, and sometimes introduce "changes" to DNA (evolution of DNA). It seems as though these walker molecules are an attempt to re-invent the wheel, or is it?

In this case, I suppose the walkers were built because a. we still don't fully understand enzymes b. we can't build our own controllable/programmable enzymes c. the walkers can do things the enzymes can't, like be non-intrusive (non-destructive), carry loads, attach and separate themselves, etc..

Now, let's talk about computational power of the walkers. Let's assume for a minute that they had advanced the walkers enough to carry a load (which they are working on). Can the walkers be programmed to release the load at a specific point? If two walkers "meet", can they join forces and combine their loads if the conditions are favorable? Would this make a single, larger walker that's twice as powerful? Ok, let's assume for a minute that we use some of the "self-assembling" nanotech that has advanced in the area of crystalline structures, and applied to the walkers.

Then we'd have to program each walker with a specific set of identifying codes (business rules) that say when the walker can merge with another, when it can't, when and what it can combine or mix it's loads with, and when it shouldn't. The end result is a tiny bit of "self-intelligence" (using the term intelligence very very loosely here). Now, if we can program and encode walkers to interact, they may actually build or self-assemble something we've never seen before.

They may actually perform "load combinations" or chemical experiments that we can't create in the lab, they may actually create new substances that we haven't seen before - pending the load itself can actually be combined to form specific results (that reaction is bound by the laws of chemistry and physics). Remember, this entry is hypothetical.

What would the combined walkers create? How would they structurally "merge" or self-assemble? How much load could combined walkers carry? My cousin, Adam Linstedt is a microbiologist professor at Carnegie Mellon University, he's suggested to me that when conditions are right, the man-made atoms (nanotech) can actually be separated from the combined chemicals. In other words, he said: making conditions favorable to binding, allows the "walkers" to bind to the target molecules, making conditions unfavorable can have the chemicals self-separate from their target molecules.

So what about the power of computing at this level? What can be said about this advancement? I believe that when this get's far enough, the walkers will indeed be encoded, and instructed to carry different loads. I also predict that self-assembly of specific kinds of walkers is inevitable and only a matter of time. The self-assembly is an interesting point when it comes to modeling.

What if we can construct models with functions that understand the context of what they contain, and what they can combine with. In other words: pair form with function, mimic the nanotech industry - we might discover new modeling and new computational models we hadn't thought of before. Self-assembling processes? Yep.

Just a thought anyhow. Cheers for now.

Share: