Rosetta and Folding

[Tlkki]

If i understand this correctly, both calculate protein refolding. Why have 2 separate projects for that?

 

[Kinslayer777]

hmm...I found a good site earlier describing somewhat objectively the difference between the primary distrubutive programs. Unfortunetly i can't find it anymore. Hiding from me at google

I believe, just roughly, that Rosetta has more to do with predicting, while folding has to do with, erm, not predicting-folds but seeing how tehy fold.. Very vague eh...


You know what. This is pretty pointless- sorry.

 

[Philippart]

http://en.wikipedia.org/wiki/Folding%40home#Similar_projects

the answer is in that paragraph on wikipedia!

 

[BlackMountainCow]

As far as I know, Rosetta and F@H both fold proteins in general. But the focus is different. Rosetta look for better methods to predict the structure of proteins - they try to improve the prediction method. F@H actually folds proteins with their methods and uses them. Of course, both always try to improve their algorithms and their methods. So it's a bit them same what they do and yet different. Now take the already finished Predictor@Home and the still runing TANPAKU project (Japanese folding project) and it gets even more confusing.

 

[BespinReactorShaft]

Originally posted by: BlackMountainCow
Rosetta look for better methods to predict the structure of proteins - they try to improve the prediction method. F@H actually folds proteins with their methods and uses them. Of course, both always try to improve their algorithms and their methods.

Err.. so basically in theory F@H should benefit from the better folding techniques discovered by Rosetta? Otherwise what synergies exist between them? Seems a bit of a waste if there aren't any...

 

[BlackMountainCow]

I guess the synergy is that whatever the Rosetta group discovers about better prediction methods, is published, subject to tests and analysis, and can thus be used by the F@H as well. On the other hand, whatever shape of a protein the F@H guys observe, the Rosetta guys afterwards can use to test their prediction methods and if they would have predicted the same structure.

 

[Philippart]

as far as I understood it:
folding@home focuses mainly on the misfolding of proteins and how to cure it, by folding it right or something. Rosetta@home fucuses on the design of new proteins and the folding of old proteins to cure diseases.

In other words: "Rosetta tries to build an army to fight the disease as a whole instead of repairing the protein which is responsible for it"

I e-mailed the chief project scientist of TANPAKU some time ago about the differences betwwen Rosetta and Tanpaku.

here's the answer:

Our final goal is to predict protein structures.
Rosetta@home has the same purpose.
The main difference between two projects is approach or method for prediction.

The method of Rosetta@home is one of the most reliable prediction method now.
However, the information obtained from the protein structures determined previously is used.

On the other hand, TANPAKU does not use the structural information. We just use information of amino acid sequence alone.
But it is very difficult task.
So, current aim of TANPAKU is development and improvement of our prediction method that suits the distributed computing.

 

[Tlkki]

Hmm, its good to hear that so many think they are not just two separate universities trying to compete against each other. That would suck.

 

[imaheadcase]

Now they should just all get together and make the program do all 3 of the techniques at the same time so they would not have to have 3 projects for it kill 3 birds with 1 stone sorta speak. Make it simpler

 

[GLeeM]

Originally posted by: Tlkki
Hmm, its good to hear that so many think they are not just two separate universities trying to compete against each other. That would suck.

In a sense they are competeing to find cures for diseases, but they share findings so the cures can be found sooner.

There is a quote from F@H leader here that I recently posted in this forum.

 

[kmmatney]

The big difference to me is that one uses BOINC and the other doesn't...