2 replies [Last post]
S0ckrates's picture
User offline. Last seen 20 hours 17 min ago. Offline
Joined: 05/19/2017
Groups: None

I've noticed something in my design puzzles lately especially since I've been looking at the designs from the partition tournament, and it's that I consistently use less residues than other players. In fact, I typically don't add any additional residues at all for my designs! But I've noticed that the added possibilities for stability shielding and bigger structures seem to outweigh the point penalties, and that other players typically add considerable amounts of residues to their designs.

Does my playstyle need updating? Should I be tacking on the additional residues in favor of a score bonus that I haven't tapped into thus far because of the mild score penalty discouraging it? Or should I still try to challenge the novel ingenuity of smaller residue count designs?

bkoep's picture
User offline. Last seen 6 hours 36 min ago. Offline
Joined: 11/15/2012
Groups: None
Good question!

Unfortunately, I can't offer any advice on your playstyle—hopefully other players can chime in. But I can say that, from the scientist's point of view, we don't intend for you to spend a lot of time thinking about the number of residues in your designs. If residue additions are outweighing the point penalties, then it's probably time we recalibrate the penalties.

Before we added the Residue Count objective, all design puzzles were fixed-length (say 70 residues, for example). The problem was that many players built designs that really only needed 67 residues, or 63 residues, etc.; the extra residues were left hanging off the end of the protein, or else were making sub-optimal interactions (in some cases, they might even destabilize the intended design).

The purpose of the Residue Count objective is to give players more flexibility in the number of residues they have to work with. We want players to insert or delete residues in order to improve their design; we don't want players adjusting their designs in order to accommodate an arbitrary residue count.

A quick glance at the most recent puzzles tells me that the top several designs usually (but not always) use the maximum allowed residues. We'll take a closer look at the residue insertion penalties to see if they need adjusting.

Joined: 09/24/2012
Groups: Go Science
Penalty policy

You are in line with the spirit of the flexibility option.
Personally, I only add no or few residues in order to make a better design, hopefully more stable.
But I noticed that I'm not well ranked in flexible design puzzles as compared with former "fix length" design puzzles. May be because the penalty is not high enough for adding residues (compared to the points gained with each additional residue)?

I would then encourage Foldit to give a higher penalty for added residues.

However, it's quite already high with about 32 pts/ added residue. I ever calculated (on old puzzles several years ago) that the mean additional point by residue was about 20pts. 20 should then be the right penalty ...

Options could be:
-rising penalty for each new residue (e.g. 32pt for n+1, 33p for n+2 etc).
-limit the number of additions to 10 (then you start with n+5 and you add or delete residues for stability).

Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons