Tools for ED puzzles - summary

Case number:845813-993956
Topic:Game: Tools
Opened by:Rav3n_pl
Status:Open
Type:Suggestion
Opened on:Friday, November 23, 2012 - 19:19
Last modified:Thursday, January 10, 2013 - 17:05

After hour of fighting with blind ED puzzle I decided to push out my frustration.
WE NEED MORE TOOLS to play denovo ED puzzles.
1. anchor view port to any point, not only segment. I wish to look on ED in different angles and I cant just rotate it around. When I pull protein out center of rotation is moving too
2. my old idea of alignment editor. It would be 1000% easier to draw shapes on ED how we see them then thread protein into that shape. I wish to draw helices, sheets, then add loops and thread protein into it.
3. editable ED config sliders. I wish to use only alpha and threshold, rest can be hidden after I set colors/shapes I prefer. Less part of screen would be covered.
4. Pin/lock in space tool. When we are sure about some fragment we want to lock it totally. any wiggle, banding or whatever should not move locked part.
5. force free/no CI mode/move frozen/locked parts. When I pull one part it should move exactly in place I pull. when I want pull helice it shpudl stay intact only loops should bend. It should not jump aside or bounce back when hit another part. When I pull sheet it suppose to stay flat and straight, not bending in strange ways

(Fri, 11/23/2012 - 19:19  |  13 comments)


jeff101's picture
User offline. Last seen 1 hour 32 min ago. Offline
Joined: 04/20/2012
Groups: Go Science

http://fold.it/portal/node/989259 is about the Alignment Editor.

I think hoops would be helpful on these puzzles too (see http://fold.it/portal/node/992670 for details).

bertro's picture
User offline. Last seen 3 weeks 3 days ago. Offline
Joined: 05/02/2011
Groups: Beta Folders

I would like to be able to:

- hide all bands (on/off checkbox in the density window)
- hide the protein and shift the mouse control focus to the density (maybe with a ctrl/alt key combination)

while looking at the density cloud.

beta_helix's picture
User offline. Last seen 23 hours 32 min ago. Offline
Joined: 05/09/2008
Groups: None

assigning...

Joined: 09/21/2011
Groups: Void Crushers
Susume's picture
User offline. Last seen 3 days 1 hour ago. Offline
Joined: 10/02/2011

Another wish: some way to make notes of where we have found features of the cloud, such as a helix or a tryptophan. We need a way to mark and reference spots in the cloud before we have figured out which pieces of protein to put there, or while we move pieces of the protein around. Two different ideas of how to do this:

1. Visible xyz axes that stay aligned to the cloud (not the protein) and a toggle on the ED menu to hide/show them. Color gradation (for direction) and course tick marks on them would be great. This would help with not getting lost when rotating the thing, and allow us to make notes (outside the game) of roughly where in space we have found features.

2. Notes attached to the cloud.

Joined: 09/21/2011

we discussed this in chat. maybe a stack of aas and a move tool so we can place these "aa markers" whe we think they go in the ed. plus instead of draw as rav suggests. how about a stack of 1 segment long sticks we can use move tool and place in the ed to try and build a mock up of backbone.

marie_s's picture
User offline. Last seen 1 year 46 weeks ago. Offline
Joined: 05/18/2008
Groups: None

I dont understand the score at all and I need usual relative scoring to play.
You should be allowed to play this puzzle as another puzzle and use the cloud as a guide without interference with the score.

Joined: 03/18/2012
Groups: Gargleblasters

My feeling exactly. In earlier ED puzzles, you could pull the protein out of the cloud and operate on it as a normal puzzle. In the latest ED puzzles, with the ED cloud being tiled but only one tile drawn, you're always under the influence of an additional force of nature that gradually pulls things in peculiar shapes. I now understand that the cloud really isn't an additional force of nature operating on the protein, it's just a measurement, and an imprecise one as well. So this imprecise cloud that operates as a force field and promotes unnatural solutions. I think we should be able to turn off the force field. Turning it on to get the "cloud score" would be a condition to credit the score.

Joined: 11/07/2008
Groups: None

First off, I'm completely new, just finished the tutorials, tried the beginner puzzles for a bit, and then plunged into the advanced puzzles. So I'm probably not qualified to make suggestions, but I'd still like to try:
>>2. my old idea of alignment editor. It would be 1000% easier to draw shapes on ED how we see them then >>thread protein into that shape. I wish to draw helices, sheets, then add loops and thread protein into it.

Came here after trying to do a DeNovo to see if someone already suggested this, and happy to find there's non-newbies wanting the same thing. Though I think I sought a solution in a different direction for the same problem. The editor sounds really nice. My idea was two parts: "lift restrictions on alignment tool" and "structure tool on sequence".

"Lift restrictions": What I'd like is having the entire available database of proteins with known structure (possibly supplemented by one's own solutions to previous puzzles) available for partial threading, regardless of what puzzle you're on. Even if there's no obvious evolutionary link, 3-10 residue parts with identical sequence in different proteins might well have similar structure, if only because that happens to be a very low-energy configuration for that particular part-sequence.

"structure tool on sequence": A sequence-based (as the alignment tool, as opposed to the "shape-based" default view) structure editor might be nice. When you start with a DeNovo protein, there's exactly no data in the shape, and it takes a lot of panning to check for parts of the sequence which might be sheets or helices and designating them as such. And letters are probably easier to recognize than sidechains sticking out at all angles, with next to no CPU usage to render.
Actually, a lot of tools might benefit from an option to "aim" them using the sequence rather than the shape, as long as there's no spatial component to the user input (so, nudge and tweak are out, but lock, cut/connect, local wriggle and maybe rubber bands between two residues could work)

Joined: 11/07/2008
Groups: None

Ehm, oops, looking back, that became a different suggestion than that of the OP. Related, but not the same. Trying to avoid duplicates, I went off-topic. Sorry.

Could everyone please ignore my earlier comment except where it concerns the OP's suggestion and could a mod maybe delete both this and that comment? I don't seem to be able to. Thank you.

alwen's picture
User offline. Last seen 4 days 7 hours ago. Offline
Joined: 10/03/2011

1. anchor view port to any point, not only segment.

YES. I would find this very useful. I have a hard time focusing on the ED cloud unless I am moving it and rotating it.

Once I start putting the protein into the cloud, the center of the point-of-view is usually not centered anywhere near the center of the ED cloud, and I can't turn the cloud easily to resolve the shapes into three dimensions in my head.

4. Pin/lock in space tool. When we are sure about some fragment we want to lock it totally. any wiggle, banding or whatever should not move locked part.

This is another I would find very useful. It is annoying to have aligned a part with the cloud so the aas sit in it perfectly, work on another section, and find the first part has moved.

5. force free/no CI mode/move frozen/locked parts. When I pull one part it should move exactly in place I pull. when I want pull helice it shpudl stay intact only loops should bend. It should not jump aside or bounce back when hit another part. When I pull sheet it suppose to stay flat and straight, not bending in strange ways

This one also! If I've built a helix and all the aas fit nicely in the cloud, I don't want the end to go straggling off and distorting because I'm trying to pull sheets into the cloud. I want the protein to stay where I put it, in the shape I put it.

steveB's picture
User offline. Last seen 2 years 48 weeks ago. Offline
Joined: 01/08/2009
Groups: Void Crushers

Absolutely 100% agree with your first point alwen. The center of rotation on ED puzzles needs to be the center of the ED cloud, and not the center of the protein. At the same time, the depth cuing in the programme which gives the white foggy effect for depth should also be centered on the ED cloud for the reasons that you point out - it's whole purpose is to help visualise in 3D.

mimi's picture
User offline. Last seen 33 weeks 1 day ago. Offline
Joined: 11/17/2008
Groups: Contenders

Can I add my bit to this suggestion.
Would it be possible to change the "Align Protein to Density" button
so that usage cycles the protein through a number of possible alignments.
Currently it seems that whenever it is used it moves it into a lower scoring
and not obviously better position - why would you ever use it?

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