Join Baker Lab scientist bkoep for an in-depth look at some Foldit player-designs from puzzle 1052: 80 Residue Symmetric Trimer Design!
In addition to highlighting some of our favorite Scientist-Shared solutions, we'll look at several lower-scoring designs to illustrate issues relating to:
Check it out and leave your questions in the comments below!
this fits with a lot of the comments on 1054, a lot of us noticed we not able to keep the ligand in a position that maintained the hydrogen bonds. it was extremely frustrating to watch your bonds be broken as your score went up.
so it was encouraging to hear that hydrogen bonds might be scored in the future.
Thanks for this video bkoep: it's great to get these insights into the pros and cons of different designs.
we should first make "good" monomer witch phobic core and proper philic outside then find way to make it bond to itself.
It would be nice if we could disable the symmetric aspects of the game while designing the monomer so that it's not so cpu intensive during the beginning stages.
The monomer part of the game would also be easier to center and rotate on the screen.
btw thanks bkoep ! I find also encouraging that lower scored solutions still can be even more interesting than high scored ones. This encourage to share to scientists. It's also great that we got this "player solution" for further evolving in puzzle 1058. Quite a chalenge for our "evolver" skills.
I agree, it would be neat to switch symmetry on and off. Unfortunately, this is isn't as simple as it sounds, and we probably won't try to implement this anytime soon.
If players are finding the symmetric chains cumbersome, they can try moving the symmetric copies apart from one another, and adjusting the view so that only the active copy is visible. The overhead for using symmetry for scoring/wiggling/shaking is not actually very large (it should run about as fast as for a monomer)—I suspect a lot of the lag results from rendering the extra symmetric residues in the display.
To be clear, hydrogen bonds are scored in Foldit, and contribute (often substantially) to the total score of a good model.
At an interface, however, Foldit favors hydrophobic side chains because they contribute to stronger binding, and hydrogen bonds simply do not contribute enough energy to counteract this bias. What we want to implement is a special bonus for hydrogen bond networks at the interface, so that we can properly encourage specific interfaces—even though they aren't as strong.
Why not have an option in the View Menu so folks can decide between showing all monomers or just the main one? If there is a way to do this already, please let us know.
Thanks!
I think the problem with this is that the active monomer would still "feel" the symmetric copies, and this might lead to some confusing behavior.
For example, imagine the frustrating scenario in which the Wiggle tool blows apart your solution for seemingly no reason—when in reality, the active monomer has severe clashes with the "invisible" copies.