5 replies [Last post]
majires's picture
User offline. Last seen 33 weeks 1 hour ago. Offline
Joined: 02/27/2015
Groups: None

Happy Holydays. I'm master's course student in proteomics and structural biology.
I'm considering writing my courses' thesis as structure prediction of DUSP6 using foldit and precision-comparison with other similar-sequenced proteins.
As I know, crystal structure of DUSP6 isn't elucidated yet.
To get help of collective intelligence, I tried to upload the contest. But, contest can't upload 100+ Amino acids protein and I don't know exactly how contest and puzzles are work.
Therefore, could someone help me to predict that structure?

sequence of human DUSP6:

Domains of DUSP6 known its structure:

Joined: 06/06/2013
Groups: Gargleblasters
suggestion for help

Contact smortier ( nickname on foldit). She is the player contact person. Your protein is very large for those of us playing on small home machines, but some number of the large machine players might be willing to play it or perhaps she can get this posted as a contest that will run longer than a handful of days by adjusting the contest rules to allow something this large.
Best of luck. This program is based off Rosetta@home. It tends to reward helices in terms of points, but if you can get a contest up maybe you can get some ideas. We have some excellent de novo folders

Joined: 06/06/2013
Groups: Gargleblasters

Majires -- of course, Foldit would need a footnote...

LociOiling's picture
User offline. Last seen 36 min 55 sec ago. Offline
Joined: 12/27/2012
a little large for Foldit

The DUSP6 sequence posted is 381 residues. That's a big protein for Foldit. The recent aflatoxin puzzles were 303 residues, and about half of those were locked, preventing the backbone and sidechain from changing.

The current puzzle 1462 has 162 residues, all unlocked. That's a large protein for Foldit. Some players have computer problems with puzzles this size.

The Foldit wiki has more information about contests: http://foldit.wikia.com/wiki/Contest

Contests don't attract much attention in Foldit.

Anyone can create a contest, but contests use one of the fixed templates listed on the wiki page. These templates have not changed in several years.

[Edit: there *is* a variable-length design contest, updated the suggestion below.]

One possible approach is to create a contest from the "Freestyle Design: Variable Length" template. After opening the contest, adjust the number of residues, and "mutate" the residues to the desired amino acids. The updated protein can then be shared with group members.

There are some problems with the contest approach. One problem is that there is no way to "lock" the contest after the initial change, so other players could still add, delete, or mutate residues. Another problem is there's no way to share the mutated contest with all Foldit players, just members of a specific group.

A better approach would be to have part of DUSP6 presented as a "de-novo" puzzle. These puzzles have a fixed set of residues, and start as an extended chain. The Foldit science team creates the puzzles, so this is only likely to happen if DUSP6 is relevant to their research.

majires's picture
User offline. Last seen 33 weeks 1 hour ago. Offline
Joined: 02/27/2015
Groups: None
Well, I found that DUSP6 is

Well, I found that DUSP6 is consist of 2 domains 1hzm(154AA)-inter amino acids(49AA)-1mkp(144AA)-final amino acids(34AA). Those structure of domains are already known, so dividing into 4 segments will makes problem easier.

Joined: 09/24/2012
Groups: Go Science

Indeed, variable length is limited to 100 residues.
You can divide your puzzle in smaller peaces and use variable length.

or you take a bigger one (500 residues) and you (set ideal helix and) freeze the remaining part. I tried here:


It's extremely heavy to play, even with a relatively powerfull computer.


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