I work in a computational biology lab so I can shed some light on it.
Folding is the use of computers to simulate biological molecular structures such as protein and nucleic acid structures. When you fold at home you are letting your computer do work for computational biology researchers, usually at stanford if you are doing the regular "Folding at home".
If you are a "boinc" folder then you use Rosetta, a different program than folding at home. This usually does work for researchers at the University of Washington. My lab uses Rosetta for its work. Rosetta has been shown in numerous benchmarks to be superior to the methods used by Stanford. There is a competition every two years called CASP where protein structure researchers benchmark their software against one another.
For some reason "Folding at home" is the popular software here. Everybody at OCN obsesses over benchmarks but whenever I bring up this topic about BOINC being superior to F@H all I hear are crickets...
I guess since folding has nothing to do with FPS nobody cares...
That and I want to one day go to Stanford as a grad student
Okay maybe I should have looked this up, but I did try and it just frustrated me. Someone please tell me what folding is and why I should try it? All I know is that it is something about diseases and proteins...but what in the hell does that have to do with my computer...
Every protein in your body is made up of subunits called amino acids. Each amino acid has specific properties based on its charge distribution, structure, etc. that allow it to interact with the environment around it. You put a bunch of these together, and these interactions allow it to "fold" into a specific shape due to these forces acting on the amino acids. Since each type of protein has the exact same amino acid sequence (i.e., alpha-hemoglobin always has the exact same amino acid sequence), every type of protein should fold in the exact same way into the exact same shape.
Of course, you can probably already guess nature doesn't like to be nice like this. Once in a while, for whatever reason, a protein misfolds. Because nature evolved to depend on proteins, and because proteins have such regularity, every process that uses/depends on proteins relies on a very specific shape. If you misfold, the shape changes and now this protein cannot be used for the process it was intended for.
By itself, it's not bad since there are a million other proteins like it that can perform the same functions, but sometimes a single misfolded protein can cause the other properly-folded proteins to become permanently misfolded. Because of this, an important biological process can no longer take place and we get many of the diseases we're familiar with (Alzheimer's, Huntington's, Muscular Dystrophy, etc).
F@H is a program that uses your PC's GPU and/or CPU to compute protein folding. By itself, protein folding is too complicated for a single computer to simulate, but when broken into many pieces and distributed (distributive computation), it becomes much more manageable.
Using the data from F@H, researchers at Stanford can figure out the mechanisms of protein folding. From there, they hope to one day understand both protein folding and misfolding, and cure diseases that arise from protein misfolding.
There has already been great success with the program. To date, there are 78 research papers published using data derived from F@H.
And that ends my procrastination for tonight
Edited by r31ncarnat3d - 2/27/11 at 11:03pm