Re: [DIYbio] Protein Unfolding Algorithms?

So this might seem outright insane given the computational requirements of protein folding, but I had a though I'm not able to locate any existing resources on, but could be the basis for an actual biotech boom (i.e. not just copying and pasting stuff, but creating things.)

The concept would be, to start: take a desired topology, stick it in the computer, have the computer generate a matching docking profile, then run a sort of protein unfolding algorithm to generate a best fitting protein for that docking profile.  Eventually add in things like active sites, inhibition sites, any special functionality (e.g. what part of the protein should flex when something is docked in some arbitrary site, which direction should it flex, should it just change shape, or open another site for access to some thing to dock in it, or eject something from another site, etc.)

I figure starting with the topology --> docking site --> protein --> DNA piece would be a really good start and would likely be pretty useful, but then adding in the other stuff would be the basis for a code --> DNA compiler (or one of the smaller units of a really powerful biological IDE which could let you build complete systems from proteins on up through organs and transmitters and such.)

Any thoughts on this welcome (especially if some or all of this already exists in an open source manner I can hack on.)  Decided to heat my house this winter by running Folding@home but the curecoins generated are still $3/day in the hole for electricity so I might as well play with other computationally intensive genomics.