Foldit participating in the WeFold CASP 10 project

[beta_helix]

For the first time in CASP's 18 year history, a handful of labs worldwide have decided to combine forces for a "Collaborative group initiative": http://www.wefold.org/

Since Foldit was the first truly global collaborative structure prediction team at CASP9, we are proud to be part of this new WeFold project.
Other groups will be sharing their templates, alignments, secondary structure and contact predictions, constraints and models with us to post as Foldit puzzles, and we will be sharing your predictions with other groups for them to score using different energy functions than Rosetta, cluster and even refine.

Any final model generated using this collaborative approach will be submitted by the WeFold CASP 10 team (denoting which methods were used at each step, Foldit being one of these) and is independent from our Foldit Group submissions (which are exclusively generated by you).

Here is a list of participants:
http://wefold.files.wordpress.com/2012/03/wefold-table.pdf
with more info on the WeFold website: http://www.wefold.org

[sncrivelli]

Many thanks to FoldIt players for participating in the WeFold experiment!

Currently, FoldIt is contributing models to two branches of the WeFold experiment: WF-FUIK and WF-FUGT.

In the WF-FUIK branch, the ensemble of models generated by FoldIt players is preprocessed and clustered to obtain groups of structures that are very similar. Representatives from each group, called medoids, are selected to make the large set of FoldIt models more manageable. George Khoury from the Floudas Lab at Princeton takes care of the preprocessing and clustering. He also ranks the set of medoids using the dDFIRE statistical potential. The models are then refined using the KoBaMIN, method developed by the Levitt Lab at Stanford.

In the WF-FUGT branch, the ensemble of preprocessed FoldIt models is ranked by the GOAP statistical potential developed by Zhou and Skolnick. The models are refined using TASSER by Zhou and Skolnick.

We’re very excited about this collaboration among different labs and are looking forward to the results!