Back to Recipes Homepage
recipe picture
Recipe: Sheet Atom Bander 1.0 -- brow42
Created by brow42 87 2553
Your rating: None Average: 4.5 (55 votes)


Name: Sheet Atom Bander 1.0 -- brow42
ID: 49008
Created on: Wed, 04/30/2014 - 22:52
Updated on: Thu, 05/01/2014 - 07:57

Bands sheets by atom. Select two sheets by selection or freezing, then run the script. Good for curved sheets.

Best For


brow42's picture
User offline. Last seen 42 weeks 6 days ago. Offline
Joined: 09/19/2011
Groups: None
For banding sheets that are curve or have a slight twist

You can bond strands easily if you straighten them first. But this leaves them stiff and perfect. You might want a curved sheet. Or, this is a refinement puzzle and the sheet is already curved. But just pulling the sheets together doesn't work, because they are a little crooked or twisted.

You need to untwist the strand a bit. But there's no tool to do that, except tweak, and that will straighten the strand. Really, the only way to do that is to band the atoms on the side. You can do this in Atom View (CTRL-SHIFT-V) but it is very tedious and confusing.

This script will try to do it for you! It's not very smart though. But when it works, it saves a lot of time. To use it, just bring two strands close together, preferably with some bonds in the middle. Then, mark the strands by either freezing them (in original interface) or selecting them (in selection interface). Then just run the script!

It will try to guess whether the strands are parallel or antiparallel, working out from the two closest segments. It will then bond the strands outward from that pair. Prolines and Glycines will be skipped.

It doesn't know which side of the strand is bonding, so it bonds both, and then deletes all the bands longer than 7 A. If the bands are too close, you may get bands between opposite sides. If your strands drift apart, eventually the script will stop banding them.

The script expects regular bonding. If your sheet is too twisted, has a cis-peptide bond (usually near a proline or glycine) or a beta bulge, you need to first change those to loops. Then you can band the regular sheets on each side of the weird residue, by selecting them in turn (along with their partner strand).

The script may mis-identify parallel and anti-parallel. This may be because the closest residues were at the ends of the strands, or you have a twist. The script may band residues offset by one. This is because your strands are shifted (they may be bonded, but the bonds are very slanted) or the closest pair was not a bonding pair. You have to align the donors and acceptors better.

If some bands look wrong, you can just delete them. It may be helpful to go into Atom View to see what the script did. You can return to Cartoon view with CTRL-SHIFT-M.

Here is what the bands look like for parallel sheets:

And for anti-parallel sheets:

This is a very experimental script and rough around the edges. Version 2.0 will probably do things differently. I'm releasing this now to help people play CASP puzzles more quickly.

Cutpoints don't count as the end of a sheet, loops do.

Joined: 05/28/2012
had an issue where it banded

had an issue where it banded every pair of atoms parralel and anti-parralel

Want to try?
Add to Cookbook!
To download recipes to your cookbook, you need to have the game client running.



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