We're almost certainly not talking about MD-scale simulation here (through the high watermark of a billion-atom simulation of Cu crystals have been achieved, and I'm pretty confident that we can put a cubic micron of a biological system into the machine, and look at 1 ns dynamics of it brute force (and much longer time spans with tricks)) within the next decade.
In principle, if you could integrate these two current complementary packages
http://www.nrcam.uchc.edu/ http://www.e-cell.org/
it would be probably up to the challenge of making a pretty good model of a minimal organism.
John Clark writes:
> Transhuman Mailing List
>
> I've been hearing some talk (the current issue of Science) of writing
> a computer program that completely simulates a bacteria, in particular
> a Mycoplasma chosen because it has only 265 genes and is the simplest known
> life form that has a metabolism. It would be a huge project but Blue Gene might
> be able to handle it. Among other things we'd need to know the 3D shape of all
> 265 proteins the genes produced but some of those are already known, obtained
> from laborious X ray diffraction experiments. If we really can write such a program
> then we can truly say we understand how at least one life form works. Anyway I
> was encouraged that respectable scientists are no longer embarrassed to talk
> about such things. There is even talk of making the first artificial cell, one that
> is even simpler, perhaps with only 180 genes or so.
>