[Chimera-users] Morphing, ligands, and secondary structure

Elaine Meng meng at cgl.ucsf.edu
Tue Aug 6 09:24:37 PDT 2019

Dear Yan,
In general, the morphing is for the protein part changing conformation.  To make the ligand fly in and out, instead of trying to put it in a morph, you would save the two positions (in the binding site, outside the binding site) using the savepos command, and then use the reset command with a frames argument to interpolate from one position to the other.  In other words, you would need to use a Chimera command script for these fancier movies that have other things (rotation, labeling, coloring, “flying”) in addition to a morph.

However, the script is fairly simple and we have lots of examples, such as in this tutorial of the combined process (ligand flies in, protein morphs):  see the last three movies and example scripts in the page showing morphing, flying, and then both in the same movie.

(the movies in this page don’t play when I use Firefox, but they do when I use Safari)

To move the ligand separately from the morph, it should be a separate model, so you could just open another copy of whatever structure that has it and hide or delete the protein part so that only the ligand is shown from that model.  Then save the ligand in the bound position with savepos, then move it to the unbound position (without moving the other models), save the unbound position with savepos. Generally you would have already calculated the morph and you would save that as well as the saved positions in a session file.

To move a model without moving other models, you can “freeze” (deactivate) the ones you don’t want to move by un-checking their “A” buttons in the Model Panel (show from Favorites menu).

You only need one stage of morphing A->B, and it only needs to contain the protein (open and closed conformations).  You only need to calculate the morph once, and then you can save it in a session.  Then you can write a script that will execute the movie content starting fom the saved session.

So you have 
(1) domain opening A->B morph (use coordset command to play morph trajectory that you already calculated)
(2) ligand flies out (reset to unbound position, previously saved)
(3) ligand flies back in (reset to bound position, previously saved)
(4) domain closing B->A morph (use coordset command to play morph trajectory backwards)

More example scripts
List of movie-related commands

Regarding the secondary structure changes, with the suggestions above you would only have one morph and you can either use ksdssp perframe or not.  Another option is to just use a spaghetti- or licorice-like ribbon that doesn’t look different for the different secondary structures (e.g. command “ribscale licorice”).

I hope this helps,
Elaine C. Meng, Ph.D.                       
UCSF Chimera(X) team
Department of Pharmaceutical Chemistry
University of California, San Francisco

> On Aug 5, 2019, at 3:49 PM, Yan Kung <ykung at brynmawr.edu> wrote:
> Hi Chimera,
> I’m a new-ish user, and I’ve found that morphing structures is usually very straightforward and easy to use. But I’m now running into an issue with morphing and ligands (and secondary structure).
> I’d like to make a single morph composed of states A-B-C-B-A (so 4 movements between 5 states). These movements are: (1) A-B: domain opens, (2) B-C: a ligand flies out, (3) C-B: the same ligand flies back in, and (4) B-A: domain closes. First problem: the ligand does not appear in the resulting morph, a problem suggested by the documentation, which I have tried (and failed) to solve using the tips I found in archived posts from this mailing list. Also, because the ligand has to fly in and out, I can’t simply superimpose the ligand from any of the individual states over the morph.
> So I tried a workaround: make each movement a separate morph. This works for movements 2 and 3 (where the ligand flies out and in). But the ligand still does not appear in movements 1 and 4 (where the ligand remains in place and the domain moves). I’m not sure why the ligand’s appearance would depend on whether it flies in/out, or whether a domain moves… But it seems to.
> Moreover, a second problem emerges that has to do with secondary structure and is difficult to explain. (Essentially: a few protein regions are on the border between helices and loops such that they appear as helices in some states but as loops in others, and vice versa. If possible, I’d like the secondary structure to remain constant across all states, or else it’s distracting. Even still, I’ve tried to embrace this and use ksdssp to show the helices/loop switching back and forth, but this is distracting as they flicker between helices and loops, each region at its own separate frame. Specifying ksdssp only in certain frames to minimize the flickering doesn’t work either, as different regions switch back and forth at different times.) The only solution I can think of that would maintain the same secondary structure throughout is to just make a single morph of all movements (and not use ksdssp). But this puts me back where I started!
> Any ideas?
> (Separately, I realize that I could make a simpler morph of just A-B-C, then record it as “roundtrip”, i.e. forward and backwards, to achieve the same thing as A-B-C-B-A. But this doesn’t solve the problems above.)
> Thanks very much,
> Yan

More information about the Chimera-users mailing list