Before they learned to put several transistors on one piece of silicon, they'd mount several separate transistors in one package to make a logic gate. So someone might have said (wrongly) that integrated circuits were not very new or useful, since the first ones only had a few transistors on one piece of silicon.

What's new about this is that he made all the corners and edges out of *one* strand of DNA. And the base sequence wasn't designed by hand. And it wasn't just double-helix: each leg was four strands criss-crossing for extra stiffness.

What's useful is that if you're building complex shapes with one long strand, you can replicate that strand with lower error rates.

Functional devices will come; give them time. Right about now, we've reached the point where we can spend more time designing the device than designing the base-pairing and doing the test-tube work. Don't be too impatient. This field is actually moving quite fast.

Chris

I haven't read the Nature article yet, but this reminds me of Ned Seeman's work (see, for example, his review article "DNA NANOTECHNOLOGY: Novel DNA Constructions", Annu. Rev. Biophys. Biomol. Struct. 1998 27:225-48). Ned has been publishing about this sort of thing at least since 1982, and he's designed and built DNA cubes, Borromean rings, knots, etc.

As far as I can see, nothing at all has come of this. It's cute, yes indeed, but somebody needs to take the next step and actually do something with one of these structures, rather than designing yet another Platonic or Archimedean polyhedron made from DNA. So I think heralding this as a "breakthrough" (Roland's term) is hyperbole. I see no evidence to date for Scripps's claim that such structures might "someday guide the assembly of nanoscale circuits that extend computing performance beyond the limits set by silicon integrated circuit technology". Does anyone know of an example where this kind of DNA structure was used to to guide the assembly of some functional nanomechanical or nanoelectronic device?