Grid mode and multiple viewports
Hi, I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol. A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video (https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC) demos the feature (in an old version of Houdini). Thanks, Roden -- This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email.
Hi Roden, Somebody else would have to answer the developer part (since I don't program), but I can confirm your first issue: Sorry no, tile just changes model placement within a single 3D scene. The models can't have different zooms, although you can put some models farther away along Z, which would make them look smaller in perspective. The side view is kind of like another viewport to the same 3D scene as in the main graphics window, but you don't have control over its orientation. <https://rbvi.ucsf.edu/chimerax/docs/user/tools/sideview.html> Regards, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 5:52 AM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Hi,
I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol.
A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video (https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC) demos the feature (in an old version of Houdini).
Thanks, Roden
Hi Elaine, Many thanks! I see now for the first issue. My description "have the zoom independent for each model" might be a bit confusing. Using the rotation as an analogy, after entering `tile` mode, the default rotation is governed by `mousemode left "rotate independent"`--I would like to have a zoom mode like this one; if one changes the rotation by 'mousemode left "rotate"', then the rotation is global--this is the current zoom under the tile mode. In other words, having an independent but synchronized zooming would be great. Knowing the possibilities and potential solutions from a development perspective would be good. I might develop a multiple-viewport feature if some initial guidance can be given. Best, Roden On Sun, Jan 26, 2025 at 11:28 PM Elaine Meng <meng@cgl.ucsf.edu> wrote:
Hi Roden, Somebody else would have to answer the developer part (since I don't program), but I can confirm your first issue:
Sorry no, tile just changes model placement within a single 3D scene. The models can't have different zooms, although you can put some models farther away along Z, which would make them look smaller in perspective.
The side view is kind of like another viewport to the same 3D scene as in the main graphics window, but you don't have control over its orientation. < https://urldefense.com/v3/__https://rbvi.ucsf.edu/chimerax/docs/user/tools/s...
Regards, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 5:52 AM, Roden Deng Luo via ChimeraX-users < chimerax-users@cgl.ucsf.edu> wrote:
Hi,
I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol.
A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video ( https://urldefense.com/v3/__https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC... ) demos the feature (in an old version of Houdini).
Thanks, Roden
-- This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email.
Hi Roden, Again I have to leave the developer answer to the others, but here's another idea: The zoom probably looks weird in the tiled situation because you are using perspective, and there is only one point of perspective. If you use orthographic projection instead, it may be more like you expect, even though there is only one zoom that applies to the whole scene. E.g. try camera ortho <https://rbvi.ucsf.edu/chimerax/docs/user/commands/camera.html> To go back to perspective, you would use camera mono I hope this helps, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 11:51 PM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Hi Elaine,
Many thanks! I see now for the first issue.
My description "have the zoom independent for each model" might be a bit confusing. Using the rotation as an analogy, after entering `tile` mode, the default rotation is governed by `mousemode left "rotate independent"`--I would like to have a zoom mode like this one; if one changes the rotation by 'mousemode left "rotate"', then the rotation is global--this is the current zoom under the tile mode. In other words, having an independent but synchronized zooming would be great.
Knowing the possibilities and potential solutions from a development perspective would be good. I might develop a multiple-viewport feature if some initial guidance can be given.
Best, Roden
On Sun, Jan 26, 2025 at 11:28 PM Elaine Meng <meng@cgl.ucsf.edu> wrote: Hi Roden, Somebody else would have to answer the developer part (since I don't program), but I can confirm your first issue:
Sorry no, tile just changes model placement within a single 3D scene. The models can't have different zooms, although you can put some models farther away along Z, which would make them look smaller in perspective.
The side view is kind of like another viewport to the same 3D scene as in the main graphics window, but you don't have control over its orientation. <https://urldefense.com/v3/__https://rbvi.ucsf.edu/chimerax/docs/user/tools/s... >
Regards, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 5:52 AM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Hi,
I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol.
A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video (https://urldefense.com/v3/__https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC... ) demos the feature (in an old version of Houdini).
Thanks, Roden
This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email._______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
HI Roden, I wanted to chime in on the second question in your first email. Yes, you can have multiple viewports, and they can have their own model trees. The segmentations bundle creates its own view, camera, and tracks its own drawings. It uses an orthographic camera to show the user a 2D slice of 3D volumes and pretends that what’s shown is true 2D (this turned out not to be the best possible solution, and it would be good to refactor when there’s time, but I digress). This is the file where my “2D” viewports are defined: https://github.com/RBVI/ChimeraX/blob/develop/src/bundles/segmentations/src/... Let me know if you have more questions and I can help you with the problem you’re trying to solve. — Zach
On 27 Jan 2025, at 09:24, Elaine Meng via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Hi Roden, Again I have to leave the developer answer to the others, but here's another idea:
The zoom probably looks weird in the tiled situation because you are using perspective, and there is only one point of perspective. If you use orthographic projection instead, it may be more like you expect, even though there is only one zoom that applies to the whole scene. E.g. try
camera ortho
<https://rbvi.ucsf.edu/chimerax/docs/user/commands/camera.html>
To go back to perspective, you would use
camera mono
I hope this helps, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 11:51 PM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Hi Elaine,
Many thanks! I see now for the first issue.
My description "have the zoom independent for each model" might be a bit confusing. Using the rotation as an analogy, after entering `tile` mode, the default rotation is governed by `mousemode left "rotate independent"`--I would like to have a zoom mode like this one; if one changes the rotation by 'mousemode left "rotate"', then the rotation is global--this is the current zoom under the tile mode. In other words, having an independent but synchronized zooming would be great.
Knowing the possibilities and potential solutions from a development perspective would be good. I might develop a multiple-viewport feature if some initial guidance can be given.
Best, Roden
On Sun, Jan 26, 2025 at 11:28 PM Elaine Meng <meng@cgl.ucsf.edu> wrote: Hi Roden, Somebody else would have to answer the developer part (since I don't program), but I can confirm your first issue:
Sorry no, tile just changes model placement within a single 3D scene. The models can't have different zooms, although you can put some models farther away along Z, which would make them look smaller in perspective.
The side view is kind of like another viewport to the same 3D scene as in the main graphics window, but you don't have control over its orientation. <https://urldefense.com/v3/__https://rbvi.ucsf.edu/chimerax/docs/user/tools/s... >
Regards, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 5:52 AM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Hi,
I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol.
A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video (https://urldefense.com/v3/__https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC... ) demos the feature (in an old version of Houdini).
Thanks, Roden
This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email._______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
_______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
Hi Roden, There isn't "zoom independent" mouse mode. But you could make one. Keep in mind that the atom coordinates opened in ChimeraX are in Angstroms and zooming does not change the coordinates, it only changes the positions of the camera. For a "zoom independent" mode the approach would be to move all the models closer to each other and the camera closer so the models each take more space in the graphics window while their center positions stay (approximately) the same. ChimeraX has some very specialized additional 3d graphics panes, Side View and segmentation orthogonal views. If you have in mind adding general extra graphics views that is a huge amount of work that would make the program substantially more complicated for the user since all the commands that change camera and view settings would then need to be able to specify which graphics pane you want to change. To appreciate the complexity, there are 400 occurrences in the ChimeraX code of "main_view" which is the Python attribute for the main graphics pane. Tom
On Jan 26, 2025, at 11:51 PM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Hi Elaine,
Many thanks! I see now for the first issue.
My description "have the zoom independent for each model" might be a bit confusing. Using the rotation as an analogy, after entering `tile` mode, the default rotation is governed by `mousemode left "rotate independent"`--I would like to have a zoom mode like this one; if one changes the rotation by 'mousemode left "rotate"', then the rotation is global--this is the current zoom under the tile mode. In other words, having an independent but synchronized zooming would be great.
Knowing the possibilities and potential solutions from a development perspective would be good. I might develop a multiple-viewport feature if some initial guidance can be given.
Best, Roden
On Sun, Jan 26, 2025 at 11:28 PM Elaine Meng <meng@cgl.ucsf.edu <mailto:meng@cgl.ucsf.edu>> wrote:
Hi Roden, Somebody else would have to answer the developer part (since I don't program), but I can confirm your first issue:
Sorry no, tile just changes model placement within a single 3D scene. The models can't have different zooms, although you can put some models farther away along Z, which would make them look smaller in perspective.
The side view is kind of like another viewport to the same 3D scene as in the main graphics window, but you don't have control over its orientation. <https://urldefense.com/v3/__https://rbvi.ucsf.edu/chimerax/docs/user/tools/s... >
Regards, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 5:52 AM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu <mailto:chimerax-users@cgl.ucsf.edu>> wrote:
Hi,
I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol.
A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video (https://urldefense.com/v3/__https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC... ) demos the feature (in an old version of Houdini).
Thanks, Roden
This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email._______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
Hi Elaine, Zach, and Tom, Very useful and informative. Thanks very much! I am integrating a whole AI-driven protein design workflow (RFdiffusion, ProteinMPNN, AlphaFold2 or AlphaFold3, Rosetta Delta Delta G, etc.) and planning to use ChimeraX as a user interface (the other candidate is molstar). I am still in the assessing phase to see if I really need an extra 3D viewport and to make tradeoffs. I am planning to check out specialized additional 3d graphics panes, by learning the pseudo-2D viewports in the segmentations bundle. If there are other examples, please kindly let me know. I will keep in mind to make it as simple for the user as possible. Best, Roden On Mon, Jan 27, 2025 at 11:53 PM Tom Goddard <goddard@sonic.net> wrote:
Hi Roden,
There isn't "zoom independent" mouse mode. But you could make one. Keep in mind that the atom coordinates opened in ChimeraX are in Angstroms and zooming does not change the coordinates, it only changes the positions of the camera. For a "zoom independent" mode the approach would be to move all the models closer to each other and the camera closer so the models each take more space in the graphics window while their center positions stay (approximately) the same.
ChimeraX has some very specialized additional 3d graphics panes, Side View and segmentation orthogonal views. If you have in mind adding general extra graphics views that is a huge amount of work that would make the program substantially more complicated for the user since all the commands that change camera and view settings would then need to be able to specify which graphics pane you want to change. To appreciate the complexity, there are 400 occurrences in the ChimeraX code of "main_view" which is the Python attribute for the main graphics pane.
Tom
On Jan 26, 2025, at 11:51 PM, Roden Deng Luo via ChimeraX-users < chimerax-users@cgl.ucsf.edu> wrote:
Hi Elaine,
Many thanks! I see now for the first issue.
My description "have the zoom independent for each model" might be a bit confusing. Using the rotation as an analogy, after entering `tile` mode, the default rotation is governed by `mousemode left "rotate independent"`--I would like to have a zoom mode like this one; if one changes the rotation by 'mousemode left "rotate"', then the rotation is global--this is the current zoom under the tile mode. In other words, having an independent but synchronized zooming would be great.
Knowing the possibilities and potential solutions from a development perspective would be good. I might develop a multiple-viewport feature if some initial guidance can be given.
Best, Roden
On Sun, Jan 26, 2025 at 11:28 PM Elaine Meng <meng@cgl.ucsf.edu> wrote:
Hi Roden, Somebody else would have to answer the developer part (since I don't program), but I can confirm your first issue:
Sorry no, tile just changes model placement within a single 3D scene. The models can't have different zooms, although you can put some models farther away along Z, which would make them look smaller in perspective.
The side view is kind of like another viewport to the same 3D scene as in the main graphics window, but you don't have control over its orientation. < https://urldefense.com/v3/__https://rbvi.ucsf.edu/chimerax/docs/user/tools/s...
Regards, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 5:52 AM, Roden Deng Luo via ChimeraX-users < chimerax-users@cgl.ucsf.edu> wrote:
Hi,
I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol.
A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video ( https://urldefense.com/v3/__https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC... ) demos the feature (in an old version of Houdini).
Thanks, Roden
------------------------------ This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email._______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/ <https://urldefense.com/v3/__https://mail.cgl.ucsf.edu/mailman/archives/list/...>
-- This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email.
Hi Roden, The Side View panel (menu Tools / General / Side View) is the only example of an extra 3D graphics pane in ChimeraX. The Segmentations panel shows 2D slices, not 3D views (although they are rendered with the 3D rendering code, but don't permit rotation). The Side View code is here https://github.com/RBVI/ChimeraX/blob/develop/src/bundles/sideview/src/tool.... For protein design you might be interested in the recently published ESM3 large language model which allows specifying structure constraints say in an active site and building out a full protein that produces those. https://www.evolutionaryscale.ai/blog/esm3-release The ESM3 Science paper a few weeks ago demonstrated designing a new GFP (green fluorescent protein) that is distant in sequence from known GFPs. ESM3 is commercial, they offer some open version but that is a much reduced version of what they used in the paper. Unfortunately much of the best protein structure machine learning such as AlphaFold 3 has restrictive licenses, not surprising since it takes hundreds of thousands of dollars of GPU time to develop and train these networks. Tom
On Jan 31, 2025, at 12:11 AM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:
Hi Elaine, Zach, and Tom,
Very useful and informative. Thanks very much!
I am integrating a whole AI-driven protein design workflow (RFdiffusion, ProteinMPNN, AlphaFold2 or AlphaFold3, Rosetta Delta Delta G, etc.) and planning to use ChimeraX as a user interface (the other candidate is molstar). I am still in the assessing phase to see if I really need an extra 3D viewport and to make tradeoffs. I am planning to check out specialized additional 3d graphics panes, by learning the pseudo-2D viewports in the segmentations bundle. If there are other examples, please kindly let me know. I will keep in mind to make it as simple for the user as possible.
Best, Roden
On Mon, Jan 27, 2025 at 11:53 PM Tom Goddard <goddard@sonic.net <mailto:goddard@sonic.net>> wrote:
Hi Roden,
There isn't "zoom independent" mouse mode. But you could make one. Keep in mind that the atom coordinates opened in ChimeraX are in Angstroms and zooming does not change the coordinates, it only changes the positions of the camera. For a "zoom independent" mode the approach would be to move all the models closer to each other and the camera closer so the models each take more space in the graphics window while their center positions stay (approximately) the same.
ChimeraX has some very specialized additional 3d graphics panes, Side View and segmentation orthogonal views. If you have in mind adding general extra graphics views that is a huge amount of work that would make the program substantially more complicated for the user since all the commands that change camera and view settings would then need to be able to specify which graphics pane you want to change. To appreciate the complexity, there are 400 occurrences in the ChimeraX code of "main_view" which is the Python attribute for the main graphics pane.
Tom
On Jan 26, 2025, at 11:51 PM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu <mailto:chimerax-users@cgl.ucsf.edu>> wrote:
Hi Elaine,
Many thanks! I see now for the first issue.
My description "have the zoom independent for each model" might be a bit confusing. Using the rotation as an analogy, after entering `tile` mode, the default rotation is governed by `mousemode left "rotate independent"`--I would like to have a zoom mode like this one; if one changes the rotation by 'mousemode left "rotate"', then the rotation is global--this is the current zoom under the tile mode. In other words, having an independent but synchronized zooming would be great.
Knowing the possibilities and potential solutions from a development perspective would be good. I might develop a multiple-viewport feature if some initial guidance can be given.
Best, Roden
On Sun, Jan 26, 2025 at 11:28 PM Elaine Meng <meng@cgl.ucsf.edu <mailto:meng@cgl.ucsf.edu>> wrote:
Hi Roden, Somebody else would have to answer the developer part (since I don't program), but I can confirm your first issue:
Sorry no, tile just changes model placement within a single 3D scene. The models can't have different zooms, although you can put some models farther away along Z, which would make them look smaller in perspective.
The side view is kind of like another viewport to the same 3D scene as in the main graphics window, but you don't have control over its orientation. <https://urldefense.com/v3/__https://rbvi.ucsf.edu/chimerax/docs/user/tools/s... >
Regards, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 5:52 AM, Roden Deng Luo via ChimeraX-users <chimerax-users@cgl.ucsf.edu <mailto:chimerax-users@cgl.ucsf.edu>> wrote:
Hi,
I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol.
A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video (https://urldefense.com/v3/__https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC... ) demos the feature (in an old version of Houdini).
Thanks, Roden
This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email._______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu <mailto:chimerax-users@cgl.ucsf.edu> To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu <mailto:chimerax-users-leave@cgl.ucsf.edu> Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/ <https://urldefense.com/v3/__https://mail.cgl.ucsf.edu/mailman/archives/list/...>
This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email._______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
Hi Tom, Many thanks! The following digresses (started from me, though..). I am happy to discuss them here or in a new thread. I had only heard of ESM3 before. I checked it out briefly. It is indeed a great achievement. However, there are two aspects hindering me from using it as a starting point. One is while I see in Section A.2.2.6 in its bioRxiv some mentioning of the interface sampling, I do not see a detailed enough (to the point one can reproduce) binder-target design workflow or results in either the paper or the code base. The other is, as you mentioned, it is a less open model, than the ones I mentioned earlier. If I understand the logic correctly, one can load the lightweight version "esm3-open" to the local GPU memory. For the stronger models, one can only interface with them on a server through API. I guess the idea is to allow the users to learn how to use the lightweight model and then transition to the full model for more powerful use cases. Currently, I have been focusing on the binder-target design as a starting point because this task involves more visualizations and interactions (my main research direction) by its nature. Now, I am able to reproduce the SHRT binder design in this Nature paper [1] published three weeks ago. The involved pipeline can all be deployed locally (but as of writing, still under some complex environments, different conda envs and docker containers). This paper builds on top of 4 other papers [2-5], all from the same group, Baker Lab. The idea is to develop a framework to abstract the workflow for protein design and develop the necessary coupling modules to make it easy to use, especially the intention specifying (or prompting) step before running the model and the generated results analysis after running the model. In the process, I will keep in mind to give the possibility of allowing other models, either open or through API, such as ESM3, to replace some of the modules in the framework. 1: https://www.nature.com/articles/s41586-024-08393-x Binder design with RFdiffusion 2: https://www.nature.com/articles/s41586-023-06415-8 RFdiffusion to design backbones 3: https://www.science.org/doi/10.1126/science.add2187 ProteinMPNN to design sequences from backbones 4: https://www.nature.com/articles/s41467-023-38328-5 Enhanced pipeline and identified threshold values for quality control 5: https://www.nature.com/articles/s41586-022-04654-9 Binder design with Rosetta Best, Roden On Fri, Jan 31, 2025 at 11:44 PM Tom Goddard <goddard@sonic.net> wrote:
Hi Roden,
The Side View panel (menu Tools / General / Side View) is the only example of an extra 3D graphics pane in ChimeraX. The Segmentations panel shows 2D slices, not 3D views (although they are rendered with the 3D rendering code, but don't permit rotation). The Side View code is here
https://github.com/RBVI/ChimeraX/blob/develop/src/bundles/sideview/src/tool.... <https://urldefense.com/v3/__https://github.com/RBVI/ChimeraX/blob/develop/sr...>
For protein design you might be interested in the recently published ESM3 large language model which allows specifying structure constraints say in an active site and building out a full protein that produces those.
https://www.evolutionaryscale.ai/blog/esm3-release <https://urldefense.com/v3/__https://www.evolutionaryscale.ai/blog/esm3-relea...>
The ESM3 Science paper a few weeks ago demonstrated designing a new GFP (green fluorescent protein) that is distant in sequence from known GFPs. ESM3 is commercial, they offer some open version but that is a much reduced version of what they used in the paper. Unfortunately much of the best protein structure machine learning such as AlphaFold 3 has restrictive licenses, not surprising since it takes hundreds of thousands of dollars of GPU time to develop and train these networks.
Tom
On Jan 31, 2025, at 12:11 AM, Roden Deng Luo via ChimeraX-users < chimerax-users@cgl.ucsf.edu> wrote:
Hi Elaine, Zach, and Tom,
Very useful and informative. Thanks very much!
I am integrating a whole AI-driven protein design workflow (RFdiffusion, ProteinMPNN, AlphaFold2 or AlphaFold3, Rosetta Delta Delta G, etc.) and planning to use ChimeraX as a user interface (the other candidate is molstar). I am still in the assessing phase to see if I really need an extra 3D viewport and to make tradeoffs. I am planning to check out specialized additional 3d graphics panes, by learning the pseudo-2D viewports in the segmentations bundle. If there are other examples, please kindly let me know. I will keep in mind to make it as simple for the user as possible.
Best, Roden
On Mon, Jan 27, 2025 at 11:53 PM Tom Goddard <goddard@sonic.net> wrote:
Hi Roden,
There isn't "zoom independent" mouse mode. But you could make one. Keep in mind that the atom coordinates opened in ChimeraX are in Angstroms and zooming does not change the coordinates, it only changes the positions of the camera. For a "zoom independent" mode the approach would be to move all the models closer to each other and the camera closer so the models each take more space in the graphics window while their center positions stay (approximately) the same.
ChimeraX has some very specialized additional 3d graphics panes, Side View and segmentation orthogonal views. If you have in mind adding general extra graphics views that is a huge amount of work that would make the program substantially more complicated for the user since all the commands that change camera and view settings would then need to be able to specify which graphics pane you want to change. To appreciate the complexity, there are 400 occurrences in the ChimeraX code of "main_view" which is the Python attribute for the main graphics pane.
Tom
On Jan 26, 2025, at 11:51 PM, Roden Deng Luo via ChimeraX-users < chimerax-users@cgl.ucsf.edu> wrote:
Hi Elaine,
Many thanks! I see now for the first issue.
My description "have the zoom independent for each model" might be a bit confusing. Using the rotation as an analogy, after entering `tile` mode, the default rotation is governed by `mousemode left "rotate independent"`--I would like to have a zoom mode like this one; if one changes the rotation by 'mousemode left "rotate"', then the rotation is global--this is the current zoom under the tile mode. In other words, having an independent but synchronized zooming would be great.
Knowing the possibilities and potential solutions from a development perspective would be good. I might develop a multiple-viewport feature if some initial guidance can be given.
Best, Roden
On Sun, Jan 26, 2025 at 11:28 PM Elaine Meng <meng@cgl.ucsf.edu> wrote:
Hi Roden, Somebody else would have to answer the developer part (since I don't program), but I can confirm your first issue:
Sorry no, tile just changes model placement within a single 3D scene. The models can't have different zooms, although you can put some models farther away along Z, which would make them look smaller in perspective.
The side view is kind of like another viewport to the same 3D scene as in the main graphics window, but you don't have control over its orientation. < https://urldefense.com/v3/__https://rbvi.ucsf.edu/chimerax/docs/user/tools/s...
Regards, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 26, 2025, at 5:52 AM, Roden Deng Luo via ChimeraX-users < chimerax-users@cgl.ucsf.edu> wrote:
Hi,
I find there is a `tile` command. After I enter the tile mode, the zoom is global. I guess `tile` is implemented by transforming the active models in the global scene explicitly. I wonder if there is a way to have the zoom independent for each model, similar to the grid mode in PyMol.
A further question, more from a developer's perspective, is whether having multiple 3D rendering viewports in ChimeraX is possible. I imagine each viewport can have its own active models in the scene and its own camera. Different cameras and other operations might be synced upon users' request. This video ( https://urldefense.com/v3/__https://youtu.be/7KZ1P-nPav8?si=Z1hiBBuRiWop2WFC... ) demos the feature (in an old version of Houdini).
Thanks, Roden
------------------------------ This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email._______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/ <https://urldefense.com/v3/__https://mail.cgl.ucsf.edu/mailman/archives/list/...>
------------------------------ This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email._______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/ <https://urldefense.com/v3/__https://mail.cgl.ucsf.edu/mailman/archives/list/...>
-- This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email.
participants (4)
-
Elaine Meng -
Roden Deng Luo -
Tom Goddard -
Zach Pearson