Could one self-assemble magnetic nanoparticles (permanent magnets to be twisted, or metal that will be pulled by external magnetic fields) into a DNA origami nanomachine, and use them for actuation?
If the position of one magnet interfered mechanically with the motion of another magnet, then you could build rod-logic, and decode a series of external field manipulations into arbitrary patterns of actuation.
Small particles for computation, larger particles for actuation (force should scale as cube of length).
Of course, rod-logic can directly sense physical/mechanical state, including some kinds of molecule binding.
Any reason to favor magnetic force over electrostatic force? I ask because pretty much all of organic chemistry depends on things happening because of electrostatic polarities rather than magnetic polarities.
That said, I don't see why not. I couldn't find any work that seemed to be directly relevant to your question, but a cursory search of the web yielded a lot of interesting stuff using magnetic nano-particles in biological systems.
In general, I like your idea. (At a meta level, I like that you are always coming up with innovative ideas, as well.) However, while I admit I have not done the detailed calculations, scaling as the cube works both ways of course. My intuition is that the net magnetic moments at scales of a few nm would be too feeble / unreliable to be practical. As Jim alluded to, the electrostatic forces would I suspect overwhelm the magnetic forces at that scale. The volatility of magnetic domains that small could also be a problem. Perhaps a quick calculation of the hysteresis of a magnetic domain at that scale compared with, say, one kT of energy would be worthwhile.
Steven C. Vetter
President, Molecular Manufacturing Enterprises, Inc.
4105 Countryview Drive
Eagan, MN 55123