Folding Forum

On the site (http://copernicus-computing.org/) it says "we observe structures 0.6Å from the native state within 30h" and "achieve sufficient sampling to predict the native state without a priori knowledge after 80-90h".

Which type and how many processors did you guys use to get those types of results?

Copernicus and FAH are not trying to solve the same problem. FAH's objective is not about predicting the native state. It's one outcome of FAH, but it's incidental to FAH's main objective which is understanding the entire folding process. Copernicus can predict the most common folding endpoint for of the Alzheimer's protein very quickly but that doesn't determine why some small percentage of proteins fold into another state. FAH needs to understand those mis-folded proteins because they are related to the causes of the disease.

Jonathan, I'm sure if you used Google to search for the paper on Copernicus you could find it and read about it. It has been my experience that scientific publications from the Pande Group have been very informative and fascinating. That would be my recommendation. This search may be a start.

F@h is focused on understanding protein folding, which as codysluder has noted and as further explained here is very much related to disease research. This is F@h's strength and what it focuses on. To a lesser degree it also performs protein structure prediction. I've read about this here and here, although there are some scientific publications that mention it.

Knowing the "native state" of an arbitrary protein is useful information. To that end, nobody should suggest that structure prediction is unimportant. There are regular contests between scientists who are working on that problem. (If you're interested, you can google for "CASP") FAH has participated in those contests but it's not FAH's primary goal, so the fact that somebody does it better/faster doesn't negate FAH's research.

There's even a computer game that lets you try to fold proteins manually, with the goal being something closer to the native state than other people have achieved.
http://www.cs.washington.edu/homes/zora ... t-2011.pdf
FAH would be interested in understanding the different thought process that the contestants used to get the better scores.

I know nothing about Copernicus but it appears to be a private company seeking patents and proprietary protection for their work rather than a University doing public funded research with the goal of placing their results in the public domain. (If I'm wrong about that, I apologize.) Some research is designed to cooperatively advance science (public research) and other research that seeks to maximize the profits of a private company (private research). Both are important, but personally, I choose to donate to publicly funded research.

By and large, scientists are both cooperative with each other and competitive with each other. Since they're solving different aspects of a bigger problem, Pande Labs and Baker Labs (Foldit and Rosetta) use each others results as the basis of new studies.

A clarification: Copernicus is an open source software from 3 academic labs (Lindahl, Kasson, Pande labs). We're releasing it at Supercomputing 2011. In a sound bite, it's "Folding@home for supercomputers", but not for contributing time to FAH as much as to take all of the scientific methodological advances in FAH's backend and bring it to other researchers.

We have a paper on our FAH web site that describes the first results. My hope is that this will lead to an even broader application of FAH's methodology. While the paper concentrates on structure prediction as a demo application, one can do a lot more with it.

Thanks for all the responses. Very interesting, even though I knew it, I am sure others learned something though, so not a waste. I am no expert (just a second year chem student) but from what I am looking up it seems to me I would need the native state, which is why I am interested in this.

To Jesse_V: I tried looking for a paper related to it but it just repeats the same thing on the home (and only) page of the website.

To bruce: if it is a private company than why is it on f@h results and at the end of the site states: "Coming soon: The first public release of Copernicus."Code = strike-though To Dr.Pande: Thank you for your response, can't wait to see it. Playing with spartan, pymol, openMM, and soon Copernicus

We'll release the paper once SC'11 is over (it's embargoed until then).

VijayPande wrote:We'll release the paper once SC'11 is over (it's embargoed until then).

Thanks. Does "SC'11" mean http://sc11.supercomputing.org/? I'm guessing it stands for "SuperComputing 2011" or am I overthinking something again?

You got it right

Jesse_V wrote:

VijayPande wrote:Thanks. Does "SC'11" mean http://sc11.supercomputing.org/? I'm guessing it stands for "SuperComputing 2011" or am I overthinking something again?

You are correct, many papers/results from events such as this are embargoed for a year.

Tobit wrote:many papers/results from events such as this are embargoed for a year.

That's not what he said.

VijayPande wrote:We'll release the paper once SC'11 is over (it's embargoed until then).

SC11 is Nov 14-17, and the paper is embargoed until it has been presented or published or something like that, so I think you should be able to get it by Nov 18.

Yes, the paper will be published by Nov 18, 2011.