9 replies [Last post]
Joined: 11/16/2007
Groups: Window Group

The way I've been playing with these has approximately been: 1) Pick out contact surfaces between secondary structure motifs using the cavity and bad-rotamer signals 2) Pull on the backbone to separate the problem surfaces from 1) 3) Use the wiggle and shake tools to settle the interfaces out 4) Once none of the inter-SS surfaces look terrible and wiggle isn't improving the score, refine: 4a) find a big clump of cavities or less favorable residues 4b) push the backbone in slightly to "remove" the cavities, or pull it out to let a side chain repack 4c) wiggle and shake. Thoughts? Recommendations? - Chris

Joined: 11/27/2007
Groups: None
Pretty similar to what I was

Pretty similar to what I was doing.  Though one thing I've been trying with some success is changing a sidechain rotamer, which results in a clash, pulling on the backbone to give it some room, and then using wiggle to settle it in.  I'm trying that when I find a rotamer I think should be packed a little better.

Joined: 11/16/2007
Groups: None

You obviously have the best technique so far, since you won the first two proteins!
Could you explain a bit more about step 1 - I'm lost in unfamiliar vocabulary, e.g., contact surfaces (they don't seem to touch ...), secondary structure motifs (little arms, wiggly ones, y's, ...?), cavity signals (yellow balls?), bad-rotamer signals (red blobs?).
BTW, how do you know all of the jargon?
Also, has anyone succeeded in twisting a 'misplaced' end of the protein into a better location? Do you just pull it apart and watch it laboriously refold, or do you try to guide it?
And most importantly, how do you get to the top so fast? Just pull apart the initial protein over and over until it takes you up to a high positive first score? or just use brute force to notch it up?
Does 'lock' do anything useful?
Great game, but would be more rewarding if I felt I were using better techniques ...

dimaio's picture
User offline. Last seen 14 years 13 weeks ago. Offline
Joined: 11/28/2007
Groups: None
Re: Strategies

I'm using a strategy similar to Chris' with some success.  I look for adjacent secondary structure elements that have a lot of cavities between them.  I try to drag one along the other, then run 3-4 iterations of "shake all" and "wiggle backbone".  If this improves the score, I'll keep the change; if not, I'll try shifting something else or shifting things in a different directions.

Additionally, aromatic (ring-containing) residue stacking seems to help a lot.  Moving the backbone and/or selecting rotamers so that the rings of tyrosines/phenylalanines stack can help scores significantly.  When the sidechains turn blue using the attractiveness coloring, you know they're set correctly.


ferzle's picture
User offline. Last seen 3 years 2 weeks ago. Offline
Joined: 12/04/2007
Groups: None
I don't have any strategies

I don't have any strategies because I am having a hard time even manipulating things.   I can see how I would want to fold the thing, but can't figure out HOW to make it fold where I want it to.  I just grab at different places and pull, and still haven't figured out how where I grab it determines which bonds/joints/whatever are going to rotate/stretch/whatever (It should be obvious that I am not a biochemist).  Is there some short tutorial I here explaining some of this stuff?

spvincent's picture
User offline. Last seen 10 hours 40 min ago. Offline
Joined: 12/07/2007
Groups: Contenders

It seems that whenever I try to use the interactive tools the score invariably decreases, and any attempts to use the techniques described above basically means starting from scratch. The only way to get anywhere seems to be to use the tools in the action tab: Shake : Wriggle All : and Wriggle Backbone, but if you're just going to try those in a random way there's not much point using an interactive tool in the first place. Still I'm quite ready to be told that I have zero talent for this game and that I should leave it to those who do: like the hypothetical teenager in Nepal who I think David Baker once mentioned.

I also don't understand the scoring system. I had the impression beforehand that it was going to use the same energy function that R@h currently uses: instead it appears to based on minimizing the volume of those orange spheres. It would be interesting to here more detail about this as it would help when playing the game.

An explanation of the colour scheme used would also be welcome.

Joined: 08/27/2018
Groups: None

bonita web

LociOiling's picture
User offline. Last seen 1 hour 18 min ago. Offline
Joined: 12/27/2012
paleo web

Wow, eva marias found a thread from before the beginning of time! We're talking the alpha test phase of Foldit if I've got it right. That makes spvincent one of the last dinosaurs.

Joined: 09/24/2012
Groups: Go Science
Beginners: don't give up !

spvincent in 2007: "Still I'm quite ready to be told that I have zero talent for this game and that I should leave it to those who do"

For those who know the achievements and creativity of spvincent … :D

jeff101's picture
User offline. Last seen 13 hours 54 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Some spvincent achievements:

#20 on Soloist Hall of Fame:
I'm way down at rank #234.

Author of many influential recipes:


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, Boehringer Ingelheim, RosettaCommons