New project starting soon Folding@home

[Hellburner]

Come on guys couldn't you do a little better on the name

http://www.stanford.edu/group/pandegroup/Cosm/

 

[Robor]

That's a *VERY* stupid name. I thought this was going to be some kind of joke laundry thread when I read it. Seems like a worthwhile project but I've got too much time invested in SETI@Home to switch myself...

Rob

 

[Engine]

This seems like a pretty cool project... and it uses the Cosm SDK!
Finally, some results out of Cosm.

Is a project like this really needed, though? I mean, isn't IBM building Blue Gene specifically for this very same purpose?

 

[JHutch]

This sounds like a very worthwhile project. I'll have to keep an eye on this, see if it lives up to its potential ...

JHutch

 

[hymy]

Yyyyyyyyyyyeeeeeeeeeeeeeeeeeeeeeeeesssssssssssssssss!!!!!!!!!!!!!!!!! Finnally a projest I can really understand. I'm gonna switch my p200mmx to to this project. IF I like it don't be suprised if I switch completely. Hehehehehe, is anyone gonna start a team?

 

[JHutch]

Well, since they don't even have the win32 client compiled yet, I'd day it's a little premature to start talking about jumping whole hog into the project ...



But, Hymy, I'll keep an eye on it, and if it doesn't just peter out, I'll be happy to throw a machine or two to their project (and start the official TA-F@H team. (Still, though, the name REALLY sucks!)

JHutch

 

[Russ]

Folding@Home? I think I saw an ad from them in Woman's World Magazine.

Russ, NCNE

 

[JHutch]

OK, Russ, and exactly WHY were you reading Woman's World Magazine?



JHutch

 

[Russ]

Lingerie ads.

Russ, NCNE

 

[Assimilator1]

Interesting project ,sucky name!

Why don't they just call it Protein folding?

 

[hymy]

Well, I've been thinking about this all day now. I'm kinda less enthusiastic. The project objectives are a bit vague. All proteins don't fold without help. There are a lotta proteins whose function is to keep proteins folded as well as help them to fold.

I guess what I'm trying to say is that those movies they have are bogus. I'll beleive that you can calculate the final folded conformation. But I don't beleive even for an instant that you can make an accurate movie of how it happens.

what I want to know is how much computational power are they gonna use to render stupid movies.

 

[Eug]

"Understanding how proteins self-assemble ("protein folding&quot is a holy grail of modern molecular biophysics."

Well, this is one project that could have impact on our lives, even within the next 5-10 years. Here's the abridged version of that site (with my editorials). I was wondering when somebody was gonna write software to do this...

In the past, using traditional techniques, people would investigate a disease, find out through statistics which DNA segments (chromosomal regions) they might be in, and then start investigating those areas for potential genes (and their proteins) which might be involved.

With this technology people are starting to do this without needing the first few steps.

One can pick random genes out of our genome, and build proteins theoretically from them. We already know the coding algorithm that the human body uses to convert DNA data to protein sequences. Just by knowing the DNA sequence we can predict the protein's amino acid sequence. (Think of DNA as the blueprint, and the protein as the finished working product.) However, because these proteins may fold any which way, we don't know which areas are actually involved in the protein's function.

Using computers, then one can predict its method of folding, and from the structure, guesstimate what the protein's function might be. From there one can predict certain diseases that may occur from alterations in these proteins.

Assuming their mathematical models are valid, this is VERY interesting work. However, it would be nice to know more about the group. The website doesn't give a lot of information. There are many groups in the world working on this stuff, but I'm not in the know of which group's work is the most valid.

The entire human genome has now been sequenced. There are now worldwide databases (eg. http://www.ncbi.nlm.nih.gov/Entrez/) full of information on gene sequences and variations on the theme both in normal DNA and DNA from diseased genes. Protein databases also now exist, but they are much less well-developed.

It seems that now (or at least in the very near future), the average Joe/Jill will be able to assist in medical research.

I will follow this with great interest.

 

[amok]

I'm with Engine on this one. Blue Gene will be able to perform at a rate of about 100x that of S@H or d.net. In the year it takes them to build Blue Gene we will have accomplished what it could do in less than 4 days...

If they would have jumped on this project two years ago it would have been worthwhile (at least until we heard about Blue Gene!). Right now it seems pointless considering a PFLOPS computer is being constructed to research in that area...

 

[Robor]

Cool... Since Engine and amok said Blue Gene will take care of this project I don't have to feel bad about "wasting" my CPU power on the search for aliens! Doing so would be a waste of our resources!

Rob