Hi Elaine I was trying to avoid splitting the model into several models though after trying a range of different combinations of commands, using the “split” command or the “surfcat” command does not allow me to easily “hide” and “re-display” the coulomb surface. I am assuming that either I manage to mangle the atom_selection or it gets mangled when I mock about with the split/join stuff however I am unable to perform the task I was aiming for. So, rather then spending several hours solving this, I just created two different models and loading them. I am trying to accomplish the desired effect right now. I do want to have the surface transparent and not hidden at some point because I want to fade the created surface into frame while spinning the model. With one issue solved I will now see if I can get through the initial recordings without issues. A big thank you for you answers and patience, feels like a personal support contact at this point! Best regards // Gustaf

On 12 Jun 2018, at 19:31, Elaine Meng <meng@cgl.ucsf.edu> wrote:

Hi Gustaf, Chimera uses some default rules as to what set of atoms is enclosed in the same surface, but you can specify different sets to enclose without using “split”. E.g.

surfcat blob #0:17.het surface blob

… where “blob” is just the name you gave it. Surface categories and “surfcat” command explained here: <http://www.rbvi.ucsf.edu/chimera/docs/UsersGuide/midas/surface.html#surfcats> <http://www.rbvi.ucsf.edu/chimera/docs/UsersGuide/midas/msms.html>

Splitting and combining creates and removes models, which may result in atoms no longer “knowing” which surface they go with, which may be the problem. I wouldn’t use that approach simply to control the display. I’d try using “surfcat” as mentioned above, or you could simply open your structure twice and delete everything except :17.het from one copy. Then hide and show parts or all of model #0 or model #1 as needed.

I don’t really understand your script. Seems like if you combine #0.1 and #0.2 and then close them, you would no longer have a model #0.1. You can see what models you have by looking in the Model Panel (open from Favorites menu). Also you can just hide/show the surface with

~surface surface blob ~surface surface #0:17.het

… rather than trying to make it 100% transparent. “~surface” does not delete the model, it just hides the surface of the specified set of atoms.

You can also hide/show the model as a whole, same as using the “S”hown checkbox in the Model Panel. It is a little trickier to hide a molecular surface but not the atomic model since they have the same model number, but the fancy command to do it is something like:

~modeldisp #0 & ~ @*

(undisplay model #0 that doesn’t have any atoms)

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

...

On Jun 12, 2018, at 1:07 AM, Gustaf Olsson <gustaf.olsson@lnu.se> wrote:

Hello Chimera users and developers, I’m back again.

So my movie rendering is progressing though I have of course run into more problems. I have one PDB file open, I need to create a surface around one molecule that is close to other molecules so running the surface command directly does not work as intended. I therefore have to split the PDB file and calculate a surface around a single molecule. All is well, however after this I run into some problems trying to clean up the script. I assume that I am defining the modelID incorrectly though I cannot figure out how.

The following works:

split #0 atoms :17.het surface probeRadius 0.5 vertexDensity 20.0 :17.het coulombic -5 red 0 white 5 blue :17.het wait combine #0.1-2 modelId 0 close true surftransparency 100 #0.1

This splits the model putting one residue (17) in a separate model, I calculate the surface without issue, color it by coulombic charges. I then have to combine the models (the single residue (17) and the rest of the residues) and close the old “residue 17” model. I then need to hide the newly created surface to display it at a later point in the movie. The script above works, however this does not seem very neat.

I have experimented with different other commands, including transparency

transp 100,s #0.1

Which also works well, as long as I split and combined the models. However, I have to combine the model for this to work!

split #0 atoms :17.het surface probeRadius 0.5 vertexDensity 20.0 :17.het coulombic -5 red 0 white 5 blue :17.het wait combine #0.1-2 modelId 0 close true surftransparency 100 #0.1

Everything but hiding the surface works.

transp 100,s #0.1

This does not work either in this case. I am assuming that I am not defining the model ID correctly for the surface and this is why I cannot hide it? I also tried using surfcat:

surfcat templ :17.het surf probeRadius 0.5 vertexDensity 20.0 templ coulombic -5 red 0 white 5 blue templ wait surftransparency 100 templ

or any of the following:

surftransparency 100 #0 surftransparency 100 :17.het surftransparency 100 #0:17.het transp 100,s #0

Hides the created surface, what am I doing wrong in defining the surface to hide it and show it again?

Thank you for any input and best regards // Gustaf

Hi Gustaf,

On Jun 13, 2018, at 12:35 AM, Gustaf Olsson <gustaf.olsson@lnu.se> wrote:

So, the commands suggested does work for creating a surface and hiding it though I am right back at not working even when using separate models. I cannot figure out what I am doing wrong here.

surfcat templ #1:17.het surf probeRadius 0.5 vertexDensity 20.0 templ; coulombic -5 red 0 white 5 blue templ ~surf templ surf templ

Running this creates a surface, colors it by charge and hides it, though now when I show the surface again it is all white. I understand that i need to re-issue the entire "surf probeRadius 0.5 vertexDensity 20.0 templ; coulombic -5 red 0 white 5 blue templ” command to show the exact same surface with desired colors again though this is not really what I was looking for. Hence, creating the surface, making it 100% transparent and then fading it in would be what I want.

The reason the surface turns white is because it has to be recalculated if you use diffferent probe radius and/or vertex density. When you first made the surface it was with nondefault probe radius and vertex density. If you use the same probe radius and vertex density when you show it again (with “surf”) it will not trigger the recalculation and you will still have the Coulombic coloring. I.e., use this to reshow the surface in order to keep the coloring: ~surf templ surf probeRadius 0.5 vertexDensity 20.0 templ … or if it’s the only surface of model #0, could hide/show with the following instead: ~modeldisp #0 & ~ @* modeldisp #0 & ~ @*

This dis not possible for me since I cannot manipulate the transparency of this surface using either “transparency” or “surfacetransparency” commands. What worked with my “broke” models before does not work at all now. These commands either do nothing or only affects 17.het structure and not the created surface. How do I specify the surface and make it transparent to be able to fade it in again?

The single molecule (17.het) is now model #1 and the only structure in that model. I run the command as stated above (using surfcat or just specifying the #1 or #1:17.het all produces the exact same result and issue) and end up with two “model #1”, the structure and the surface. I cannot modify the transparency of the surface at all, none of these commands work:

transp 100,s templ transp 100 templ transp 100 #1 transp 100 #1:17.het transp 100 :17.het surftransparency 100 templ surftransparency 100 #1 surftransparency 100 #1:17.het surftransparency 100 :17.het

Using the GUI works just as intended so obviously I am the issue and I’m obviously not using the correct syntax, so how do I do this correctly?

The issue is not the syntax ( and “surftransparency” is exactly the same as “transparency” with “,s” after the value so no need to try both) but that the Coulombic-colored surface is in a different coloring mode that does not allow transparency to be adjusted. It will work fine on surfaces that haven’t been colored to show per-vertex quantities. If you really want to fade in a surface, I recommend fading it in as a single-color surface and then showing Coulombic coloring after that. You could actually have two separate surfaces and fade in the non-Coulombic-colored one and then do a rapid switcheroo of hiding that one and showing the already-made Coulombic one. I hope this helps, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Department of Pharmaceutical Chemistry University of California, San Francisco

Hi Gustaf, I did test all my recommendations and assertions (and demonstrated to myself the inability to change transparency of the Coulombic-ESP coloring) so the behavior you report is a mystery to me. However, I did not try splitting/recombining, so that and other details may make the difference. Maybe you needed more “wait” commands to show the fading, if simply using separate command files helped, but never mind if the problem is solved. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Department of Pharmaceutical Chemistry University of California, San Francisco

On Jun 13, 2018, at 11:23 PM, Gustaf Olsson <gustaf.olsson@lnu.se> wrote:

Hello Elaine

Thank your for your reply, you have been a fantastic support and your answers do helt a lot.

I realised that the “surf” command led to a re-calculation and this was why I was having issues “re-displaying” the surface as intended. Rather than trying to figure out how to do it properly, I just went back and used the insanely complicated script I had before.

Two final thought on this matter for this discussion:

First, using the transparency functions, I was unable to “fade in” several structures when trying to run a “larger” script making the entire movie. Breaking the CMD file into several separate files results in all the “fading” action working properly without any other modifications. This is strange though since I am now getting a bit stretched for time I will not investigate this further.

Second:

...

The issue is not the syntax ( and “surftransparency” is exactly the same as “transparency” with “,s” after the value so no need to try both) but that the Coulombic-colored surface is in a different coloring mode that does not allow transparency to be adjusted. It will work fine on surfaces that haven’t been colored to show per-vertex quantities.

If you really want to fade in a surface, I recommend fading it in as a single-color surface and then showing Coulombic coloring after that. You could actually have two separate surfaces and fade in the non-Coulombic-colored one and then do a rapid switcheroo of hiding that one and showing the already-made Coulombic one.

This is harder to explain, using my chaotic split/combine and surftransparency combination I was able to create a surface, color it by charge, hiding it using surftransparency and fading it back in using the same. If this should not work then I am at a loss since this is exactly what I did and it works perfectly although the code is a horrible mess.

Once I get everything finished I’ll post the video somewhere and send you a link.

Thank you again and best regards // Gustaf

...

On 13 Jun 2018, at 19:48, Elaine Meng <meng@cgl.ucsf.edu> wrote:

Hi Gustaf,

...

On Jun 13, 2018, at 12:35 AM, Gustaf Olsson <gustaf.olsson@lnu.se> wrote:

So, the commands suggested does work for creating a surface and hiding it though I am right back at not working even when using separate models. I cannot figure out what I am doing wrong here.

surfcat templ #1:17.het surf probeRadius 0.5 vertexDensity 20.0 templ; coulombic -5 red 0 white 5 blue templ ~surf templ surf templ

Running this creates a surface, colors it by charge and hides it, though now when I show the surface again it is all white. I understand that i need to re-issue the entire "surf probeRadius 0.5 vertexDensity 20.0 templ; coulombic -5 red 0 white 5 blue templ” command to show the exact same surface with desired colors again though this is not really what I was looking for. Hence, creating the surface, making it 100% transparent and then fading it in would be what I want.

The reason the surface turns white is because it has to be recalculated if you use diffferent probe radius and/or vertex density. When you first made the surface it was with nondefault probe radius and vertex density. If you use the same probe radius and vertex density when you show it again (with “surf”) it will not trigger the recalculation and you will still have the Coulombic coloring. I.e., use this to reshow the surface in order to keep the coloring:

~surf templ surf probeRadius 0.5 vertexDensity 20.0 templ

… or if it’s the only surface of model #0, could hide/show with the following instead:

~modeldisp #0 & ~ @* modeldisp #0 & ~ @*

...

This dis not possible for me since I cannot manipulate the transparency of this surface using either “transparency” or “surfacetransparency” commands. What worked with my “broke” models before does not work at all now. These commands either do nothing or only affects 17.het structure and not the created surface. How do I specify the surface and make it transparent to be able to fade it in again?

The single molecule (17.het) is now model #1 and the only structure in that model. I run the command as stated above (using surfcat or just specifying the #1 or #1:17.het all produces the exact same result and issue) and end up with two “model #1”, the structure and the surface. I cannot modify the transparency of the surface at all, none of these commands work:

transp 100,s templ transp 100 templ transp 100 #1 transp 100 #1:17.het transp 100 :17.het surftransparency 100 templ surftransparency 100 #1 surftransparency 100 #1:17.het surftransparency 100 :17.het

Using the GUI works just as intended so obviously I am the issue and I’m obviously not using the correct syntax, so how do I do this correctly?

The issue is not the syntax ( and “surftransparency” is exactly the same as “transparency” with “,s” after the value so no need to try both) but that the Coulombic-colored surface is in a different coloring mode that does not allow transparency to be adjusted. It will work fine on surfaces that haven’t been colored to show per-vertex quantities.

If you really want to fade in a surface, I recommend fading it in as a single-color surface and then showing Coulombic coloring after that. You could actually have two separate surfaces and fade in the non-Coulombic-colored one and then do a rapid switcheroo of hiding that one and showing the already-made Coulombic one.

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

_______________________________________________ Chimera-users mailing list: Chimera-users@cgl.ucsf.edu Manage subscription: http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users