In this puzzle, you can redesign a protein that was found to bind Abeta, a peptide associated with Alzheimer's disease. See the blog post about this for more information. The Abeta protein backbone is frozen in this puzzle. The binding protein consists of two chains, in which you can mutate all but several loop residues. Each chain has a separate RMSD condition, so only slight backbone movements will be tolerated. This puzzle also includes a Residue IE filter, which monitors that all PHY, TYR, and TRP residues are scoring well. Remember, you can use the Upload for Scientists button for up to 3 designs that you want us to look at, even if they are not the best-scoring solutions. Help us design a tighter-binding protein for Abeta!
Puzzle 801 was posted with an incorrect score function.
You will be able to load in your manual saves from 801 into this puzzle. If you did not manually save a solution in puzzle 801, you can go back to 801, manually save it, and the solution should appear in your manual saves for this puzzle, 801b.
Could you let us know which part of the score function was changed? I've been wracking my brain trying to remember the name of score parameter that they added in a certain beta puzzle a while ago. Pretty sure it had to do with the sidechain reference scores, but am not sure.
The score term that changed between 801 and 801b was the 'envsmooth' score term. It is very similar to residue reference energy, but it contributes to the "Hiding" residue score in Foldit. Earlier this year we ran a few BETA puzzles with different envsmooth behavior, similar to this: http://fold.it/portal/node/994885.
The evolver criteria seems insanely high for me. I loaded up a teammates fold that has a score of 10091 points, I'm seeing "Evolver: not met 13013 points needed" in the conditions list.