"Idealize structure" button

Case number:845813-988104
Topic:Game: Tools
Opened by:Rav3n_pl
Opened on:Friday, July 30, 2010 - 00:03
Last modified:Tuesday, June 3, 2014 - 22:33

Sometimes, at start of folding very useful function is needed: idealize.
When I start folding I first make secondary structure then wiggle to bend helices and rebuild them to shape. Then I try align sheets and straighten them. Make sheets flat and straight by one button and make helices to swirl same way would be very useful. Sometimes rebuild takes looooong time to get good shape of helix, and flatten sheets sometimes is impossible.

(Fri, 07/30/2010 - 00:03  |  9 comments)

Joined: 05/13/2008
Groups: Boinc.be

I like that one! It can be a pain just to flip on sidechain in a sheet for instance...

Joined: 06/17/2010

A bump.
Is there a chance to do that?

Joined: 04/15/2012
Groups: Beta Folders

I like it. It is a good starting point if you have the right structures. Otherwise, it is a pain to do.

Joined: 12/06/2008
Groups: Contenders

Once we are allowed to make cutpoints anywhere we want, it should be easy to cut out a sheet or helix, straighten out its structure, then paste it back into place.

At least, that's what I hope for.

jeff101's picture
User offline. Last seen 9 hours 36 min ago. Offline
Joined: 04/20/2012
Groups: Go Science

I think one could make a LUA V2 recipe for this idealize function.
It could read the secondary structure ss (H E or L) for segment res1 using
ss=structure.GetSecondaryStructure(res1) and then use an appropriate combination of bnum=band.AddBetweenSegments(res1,res2,atom1,atom2) and band.SetGoalLength(bnum,dist)
to set the distance between atom1 on amino acid res1 and atom2 on amino acid res2
to the value dist.

One problem might be that right now band.AddBetweenSegments doesn't seem to work when
|res1-res2| is less than 2, which is a little odd because structure.GetDistance(res1,res2)
works fine for that setting. I should add that it would also be nice if structure.GetDistance
also had inputs for atom1 and atom2 like band.AddBetweenSegments has right now.

Another problem to overcome is how to set the handedness of helices.
Distances between atoms might not be enough because
mirror images have the same distances between vertices.

Of course, once you set all the bands, you'd have to at least wiggle the backbone.
You'd probably want to freeze the other parts of the protein during this wiggling
so that only the sections you want to idealize get wiggled.

See http://foldit.wikia.com/wiki/Foldit_Lua_Functions
for more details about the LUA V2 functions listed above.

brow42's picture
User offline. Last seen 8 weeks 2 days ago. Offline
Joined: 09/19/2011
Groups: None

I've tried this. It doesn't work nearly as well as it should. I didn't release the script because it would be too easy to make Pi Helices! It might be safe now. But it still doesn't work. Bands just aren't strong enough to move the protein, especially if the sidechains are already packed tight.

Scissors, rebuild, and constraint slider is the way to go here. There are lots of quirks and tweaks we'd like to fix in the Lua API, but we've only just gotten the improvements we have now, and a lot of that was done by placing Tlaloc literally under contract. There's no timeline about when or if they'll do that again. Figure out how to work with what you've got.

jflat06's picture
User offline. Last seen 1 hour 38 min ago. Offline
Joined: 09/29/2010
Groups: Window Group

This should be resolved once we have a slider to control the cutpoint forces.

Joined: 09/18/2009
Groups: SETI.Germany

I like the idealizing idea, too (I could be some kind of rebuild-function).
As helices can be curled rather easily, but sheets do not always react if you want to flatten them, or it takes a long time.
But Boots' / Brow42's idea seems to be the easiest way.

To straighten out sheets, you can also connect the first segment with the last segment of the sheet section with an enormous long and strong band and do some backbone manipulation (as far as I remember, this was a solution Tlaloc had shown). This sometimes works better than rebuild or straighten out via GUI.

Joined: 06/17/2010
Status: Open » Closed

Done! We have "idealize secondary structure" tool.


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