The Idealize Tool

Started by jflat06

jflat06 Staff Lv 1

Hey everyone,

I'd like to introduce a new tool that is coming to the Foldit client: The Idealize tool.

So what does this tool do? Well, proteins, there are certain degrees of freedom that can change (like the angle of a sidechain coming off the backbone), and some that really shouldn't (like the distance between two residues). You can see some examples of degrees of freedom which should not change here:


(credit The Rockafeller University)

You'll notice there are very specific values drawn into this picture that describe certain angles and distances between atoms. These values are called 'ideal' values, and in real proteins, their values do not change much.

In Foldit, sometimes these values start to get away from their ideal values, and you'll see this as a poor backbone score on a segment or two. These issues can be very hard to fix. The easiest way to introduce this problem is with cuts - if you make a cut and then try and close it at an odd angle, often times it will be quite difficult to massage out the problem.

With that, we've created the Idealize tool. This tool lets you simply set all of these degrees of freedom to their ideal values for a portion of the backbone. The tool works in the Classic interface via the right click menu, and will pop up in the selection interface when you have several residues selected. It is also available in the scripting interface with a call to IdealizeSelected().

For an example of where this tool can be used, look at the backbone below:

You may or may not be able to tell, but there are problems with this backbone. You can see what the backbone should look like in the following picture:

You can see that the blue portion should have a nice angle around 120 degrees or so. In the backbone on the top, however, it is near 180 degrees. This will result in a very poor backbone score in Foldit, and is something you want to fix.

And with the idealize tool, you can do just that! In the classic interface, right click the backbone where the issue exists, and select Idealize.

This will set the problematic degrees of freedom to their ideal values, and will make the backbone look like the second, correct picture. Note that while this tool only modifies values in the region you're working on, the changes cause one entire side of the protein to swing into a new position.

We hope this tool will be useful, and are happy to receive any feedback you have!

brow42 Lv 1

Will the SS assignment change the behavior of Idealize, or does Idealize only affect parts that don't change much, ever?

Very exciting tool, and may completely change the scoreboards!

jflat06 Staff Lv 1

Secondary structure does not affect the behavior of the tool. In general, the degrees of freedom that it sets are nearly constant for every residue-residue bond in nature.

We're hoping this tool will be very useful!

Timo van der Laan Lv 1

I used it to optimize a closed cutpoint. It works but it moves more than the selection wrecking the alignment of two sheets.

jflat06 Staff Lv 1

As mentioned in the post, while this tool only affects local degrees of freedom, this results in global changes. I've highlighted this note in the post in bold.

drumpeter18yrs9yrs Lv 1

Could the hotkey for this be changed? I don't like the fact that when I select a segment and want to bring up the view options by pressing v and it idealizes.

spmm Lv 1

This issue has to do with proline residues, which are handled uniquely by Rosetta. Because proline has to maintained its closed ring, "idealizing" this residue does not result in bond lengths, angles, and torsions that are considered "ideal" by the Rosetta score function. As a result, players should not expect to see an Ideality score of 0 for proline residues.

Additionally, because a residue's torsion angles depend on atoms of both the previous residue and the next residue (four consecutive atoms define a torsion angle), this also affects the Ideality score of those residues on either side of the proline.

In puzzle 1091, there are prolines at positions 17 and 36, which is why you are observing this discrepancy for segments 16-18 and 35-37. I expect you will see similar results around other prolines in the puzzle.

I understand that this seems nonsensical (i.e., the Idealize tool yields an imperfect Ideality score), but this is expected behavior, and is unlikely to change.

From bkoep's feedback replies

http://fold.it/portal/node/2000773#comment-30558
http://fold.it/portal/node/2000773#comment-30557

LociOiling Lv 1

http://foldit.wikia.com/wiki/Idealize

The wiki page has been updated with a detailed explanation of how idealize works in the original and selection interfaces. The nifty diagram from the top of this page is also there, along with the details on what the diagram is showing.

Update: the first diagram on this page was not displaying correctly for me earlier, but now it seems to be working correctly. It's hosted at rockefeller.edu, so maybe it was a problem on that end. At any rate, the same diagram from a different source is now available on the wiki.