The model that the CASP organizers have given us for this Refinement Target is actually a RosettaServer model (generated using Rosetta@home).
The trouble with this (as we learned in CASP9, which you can read about here: http://depts.washington.edu/bakerpg/drupal/node/435) is that this model is already minimized with Rosetta so it will be very hard to move that model out of its local energy minimum.
So for the starting model, we have given you the closest non-Rosetta server model (that is only 1.2 Angstroms RMSD away from the model chosen by the CASP organizers) so that you will be able to find high-scoring solutions that are slightly different than the ~80% correct starting structure.
In addition, we are providing you with the exact template and alignment that Rosetta used to generate that model, so you should be able to end up with the same structure as Rosetta (but without having it being heavily minimized). We also hope that by moving the alignment around you are able to increase the model accuracy from 80% to 100%.
![](http://128.95.1.228:8000/tracking_pixel.gif?events=scAdd&f=1111&wdata={"x":12, "y":11}){kind=tracking}
Here is the CASP link for this Refinement target:
http://predictioncenter.org/casp10/target.cgi?id=104&view=refinement
____________________________________________________
Here is the sequence logo predicted by the SAM server.
H = helix
E = sheet
C = loop (or coil)
The taller the letter at each position, the higher the probability of that specific secondary structure for that amino acid.
For example, the Leucines at residues 18-19, 22, 40 & 82 are highly predicted to be in a helix. However, the Leucine at residue 131 is predicted to be in a sheet.
____________________________________________________________
Note that we are giving you over 8 days to work on this one, because it's a large protein and TR688 will have to be due on the same day (they gave us less than 3 weeks to work on that one).