Simplify computing CA RMSD of an ensemble of structures.

[Tom Goddard]

Current steps for computing the CA RMSD of a set of structures are complex

​https://www.cgl.ucsf.edu/pipermail/chimera-users/2014-March/009709.html

and the question of how to do it has been raised many times on the Chimera and ChimeraX user lists. We should have a trivial command or gui way to do this commonly needed task. Should also set an attribute to allow easy coloring by CA RMSD.

Tristan addressed this in his answer to a CCP4 mailing list question:

On May 26, 2021, at 10:29 AM, Tristan Croll <tic20@…> wrote:

Thought you might be interested. Just sent him a quick-and-dirty ChimeraX script to color by CA-RMS (attached as .txt since Outlook gets grumpy with me otherwise)... but seems like it might be a neat little addition? Would use the bug report system but the server's down.

From: CCP4 bulletin board <CCP4BB@…> on behalf of Harry Powell - CCP4BB <0000193323b1e616-dmarc-request@…>
Sent: 26 May 2021 16:04
To: CCP4BB@… <CCP4BB@…>
Subject: [ccp4bb] Analysis of NMR ensembles

Hi

Given that there are plenty of people on this BB who are structural biologists rather than “just” crystallographers, I thought someone here might be able to help.

If I have a structure in the PDB (e.g. 2kv5) that is an ensemble of structures that fit the NOEs, is there a tool available that will give me some idea about the bits of the structure that do not vary much (“rigid”) and the bits that are all over the place (“flexible”)?

Would superpose or gesamt be a good tool for this? Ideally I’d like something that could add a figure to the B columns in a PDB file so I could see something in QTMG (or PyMol if forced…) or do other useful things with the information.

Harry

[Tom Goddard]

Tristan's script to color by CA RMS

[Tom Goddard]

I attached Tristan's Python script that he gave to the CCP4 user to do CA RMSD coloring. Sets bfactor to CA RMSD then colors by bfactor.

[Elaine Meng]

It is extremely easy. Just open the structure,

open 2kv5

then in the chain info table click the entry for that entity under Description.  For this example, it's "Putative uncharacterized protein RNAI" .  That opens the chain sequence and automatically associates all chains of that entity with it.  

Then use Sequence Viewer context menu (Ctrl-click on mac) Headers... Calpha RMSD, then command:

color byattr seq_rmsd

(appending any other options you want, like palette etc.)

Elaine

[Tom Goddard]

It is not easily discoverable how to do it and those steps are hard to explain because it uses non-obvious GUI interfaces. How would I do that with commands, I'd expect it might take two commands: compute CA RMSD and color although one command would be even nicer.

[Elaine Meng]

Doing it via GUI generates and executes the equivalent commands:

open 2kv5
sequence chain #1.1/A#1.2/A#1.3/A#1.4/A#1.5/A#1.6/A#1.7/A#1.8/A#1.9/A#1.10/A#1.11/A#1.12/A#1.13/A#1.14/A#1.15/A#1.16/A#1.17/A#1.18/A#1.19/A#1.20/A
sequence header rmsd show
color byattr seq_rmsd

The "sequence chain <really long atomspec>" thing is one command and works exactly the same with a more compact atomspec:

open 2kv5
sequence chain #1/A
sequence header rmsd show
color byattr seq_rmsd

[goddard@…]

Doing this simple computation by display of a sequence through the GUI is ok, but roundabout, and we can make this simpler.  I am having trouble understanding the resistance to that idea.  If Eric feels similarly that this is not worth any effort please reassign the ticket to me.

[Tristan Croll]

Would also be nice to be able to treat a multi-chain complex as a rigid body for alignment/RMSD calculation purposes. As far as I'm aware only chain-by-chain alignment is currently possible.
________________________________
From: ChimeraX <ChimeraX-bugs-admin@cgl.ucsf.edu>
Sent: 27 May 2021 03:22
To: pett@cgl.ucsf.edu <pett@cgl.ucsf.edu>; goddard@cgl.ucsf.edu <goddard@cgl.ucsf.edu>
Cc: meng@cgl.ucsf.edu <meng@cgl.ucsf.edu>; Tristan Croll <tic20@cam.ac.uk>
Subject: Re: [ChimeraX] #4690: Simplify computing CA RMSD of an ensemble of structures.

#4690: Simplify computing CA RMSD of an ensemble of structures.
-----------------------------------------+----------------------------
          Reporter:  Tom Goddard         |      Owner:  Eric Pettersen
              Type:  defect              |     Status:  assigned
          Priority:  moderate            |  Milestone:
         Component:  Structure Analysis  |    Version:
        Resolution:                      |   Keywords:
        Blocked By:                      |   Blocking:
Notify when closed:                      |   Platform:  all
           Project:  ChimeraX            |
-----------------------------------------+----------------------------

Comment (by goddard@…):

 {{{
 Doing this simple computation by display of a sequence through the GUI is
 ok, but roundabout, and we can make this simpler.  I am having trouble
 understanding the resistance to that idea.  If Eric feels similarly that
 this is not worth any effort please reassign the ticket to me.

 }}}

--
Ticket URL: <https://plato.cgl.ucsf.edu/trac/ChimeraX/ticket/4690#comment:5>
ChimeraX <http://www.rbvi.ucsf.edu/chimerax/>
ChimeraX Issue Tracker

[pett]

I guess I have some questions about the parameters of the request, both on the input side and the output side.

Input:

1) Ensemble has identical sequences?
2) Ensemble has identical missing segments?
3) Are we talking about coordsets as well, or only individual structures?

Output:

4) You get a list of residues and RMSDs in the log and/or file?

[goddard@…]

The command design would be up to you.  Of course ideally it would handle all the input cases but whether those are easy to implement is a factor.  For output I thought assigning an attribute.  Optionally output to a file or to the log would be a nice touch but I wouldn't make that the default.

This request should not have especially high priority relative to the hundreds of other feature requests we have.  But with the number of times this gets asked I'd put it at least in the middle of that pack, not at lowest priority.  But up to you of course, or me if assigned to me.