Pins Tool

Case number:845813-990659
Topic:Game: Tools
Opened by:tristanlbailey
Opened on:Saturday, October 1, 2011 - 07:33
Last modified:Wednesday, March 20, 2013 - 00:02

I would like to suggest the addition of a new tool: Pins.

Pins would operate like a cross between Bands and the Freeze tool, in that a pinned segment is NOT allowed to move from point 'A' to point 'B' (locomotion), but is allowed to rotate in place (i.e. its current position).

Note that it is possible for bands to behave in this manner, except with pins, the hassle of placing the band as close as possible to its local segment, and then increasing the band's strength, is removed. With multiple pins versus multiple bands, you are negating the chance of considerable unwanted locomotion.

It's appearance could use a lime green colour, to provide the best possible contrast for their quick location, amongst other colours currently being used. It should use a sphere for it's shape (like bands), except be slightly larger to noticeably protrude from segments.

Perhaps this tool could be implemented by using the Alt + Left Click combination, as this combination doesn't appear to be assigned to anything concurrently. Since pins would NOT serve an extended function, it is recommended that their removal should simply be confined to a simple Left Click with a mouse.

I have included a couple of pictures, to try to illustrate my suggestion.

Before user drags segment up between pins66.59 KB
After user has finished dragging upwards65.61 KB
(Sat, 10/01/2011 - 07:33  |  11 comments)

Joined: 09/22/2011
Groups: None

Updated After Drag picture to show segment rotation.

Joined: 09/22/2011
Groups: None

After having a quick chat to thomirc8, I figure that this idea could be extended to include 2D and 3D pin versions.

The 2D version would allow a segment to rotate only along the 2D axis at the point it entered. This is akin to someone attaching a piece of paper with a pin, to a cork board. The paper can rotate along a 2D axis at the point of the pin, but cannot move locomotively (ie. up, down, left, right, etc.). Another example would be someone using a curve-line tool in an image editing program. After points are created, the movement of the mouse in any direction causes the line to curve smoothly through those points (conform), before the line is finalised.

The 3D version would allow a segment to rotate freely along a 3D axis, without allowing for locomotion. This would be analogous to a rotation tool in a 3D graphics modelling program (i.e. Maya, 3ds Max, Blender, etc.), where the object, or section of an object, rotates around a point, but does NOT move locomotively from that point.

The Pins tool does NOT attempt to directly rotate certain segments, but allows for segment rotation by other tools, without moving the segments from their designated points, defined by the placing of a pin.

The best use of the pins tool that I can think of, is with the use of the Rebuild tool. Imagine you are trying to restrict parts of a protein to particular areas while using the Rebuild tool, but allowing the rest of the protein to move about freely. The Freeze tool is only useful for supposedly holding segments of a protein exactly where they are, not allowing for any motion of those particular segments at all, including rotation (currently disputed). The Rebuild tool cannot rotate these frozen segments in place; in fact, the rebuild tool is restricted to sections of segments between frozen segments, and/or ends of a protein. The pins tool, however, would allow for the rotation of pinned segments in their places, still without changing the general conformation set by the user.

The pins tool can, of course, be used in conjunction with other tools, such as Shake, Wiggle, and even Bands. I am NOT suggesting this tool to be a replacement of the bands tool, nor am I suggesting that it replace the Freeze tool.

Again, I must stress, that although bands can perform this behaviour, it requires a lot more time to position several bands in the exact position as their segments, with a good chance of still causing a small amount of locomotion, in any one particular direction, per segment. With multiple segments all moving locomotively a fraction, it could equate to considerable shift away from the user's ideal conformance.

Joined: 06/17/2010

It looks like combination of use "q" key on segment (to change rotate viewport) and locking frozen segment in space :)

Joined: 09/22/2011
Groups: None

Umm... that's not quite what I was getting at. The camera view is not necessarily changed to pivot/rotate around pinned segments, but that is something else that could be considered as an additional option under the View menu, when using pins.

I was referring to the segments themselves (and by extent, the backbone). The backbone, when it is pulled / wiggled / rebuilt, allows pinned segments in the backbone the ability to rotate/pivot/bend in place, but NOT move from their pinned location (i.e. no locomotion allowed). In other words, their angles (X, Y, Z with "3D pins") can change, just not their location.

The pins tool would be more like the bands tool, than anything else. When you have a particular structural conformation in mind, the general idea is that they can be used faster than bands, for rebuilding protein structures quicker. There's no fiddling around to make sure that bands are placed as close as possible to the centres of each segment in concern.

Note that pins are not supposed to have length or strength values, since their associated segments do not move locomotively (i.e. they are fixed in position).

steveB's picture
User offline. Last seen 2 years 3 days ago. Offline
Joined: 01/08/2009
Groups: Void Crushers

I think it's a good idea. I use "pins" (ie zero length bands) frequently and to be able to set them easily would be a great help.

Joined: 09/22/2011
Groups: None

Thanks, steveB. Yes, they could be thought of as "zero length bands".

Something else I realised recently, is that when you are working on a protein with lots of amino acids and folds, it can be difficult to use bands to "pin" or manoeuvre segments and/or sidechains to particular spots in space (rather than to other segments), as bands are set to automatically "snap" to segments and sidechains in the background. To overcome this problem, you usually have to try and change the camera view, so that there is grey/white background space to stretch the bands to.

Perhaps, as well as implementing pins, the bands tool should be reviewed in this regard. Maybe programming Foldit to use a key on the keyboard (to hold down while dragging) could be used to prevent "snapping".

Maybe this bands idea could go into another Feedback item?

wlemmon's picture
User offline. Last seen 8 years 46 weeks ago. Offline
Joined: 07/10/2010
Groups: None

Pins is a great idea.

Joined: 04/15/2012
Groups: Beta Folders

Also, if you could only focous on the parts not pinned, that would help. That way you can just work on one part at a time, maybe with a show/hide pinned sections button. Shift-Q does pretty much the same thing, but it often shows multiple sections, and when I move the protein, it shows other parts.

jflat06's picture
User offline. Last seen 17 hours 34 min ago. Offline
Joined: 09/29/2010
Groups: Window Group
Status: Open » Open

I have an idea of how this could be made to work, and I'll try to test it out. It probably wont happen for a while, though.

Joined: 04/15/2012
Groups: Beta Folders

Good to know it is on the agenda though. Thanks!

Joined: 04/15/2012
Groups: Beta Folders

Bump. This would be extremely useful on 688. I can't get the protein to stay still. I want to resolve clashes (just a short wiggle), but leave just the bonded segment where it is. Bands/freezing don't work. It still moves, even with a strength of 10 and length of 0. If the segment moves by the slightest bit, the bond is breaks.


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