#1910 closed defect (nonchimerax)
Crash with symmetry expansion
| Reported by: | Owned by: | Tom Goddard | |
|---|---|---|---|
| Priority: | normal | Milestone: | |
| Component: | Higher-Order Structure | Version: | |
| Keywords: | Cc: | Eric Pettersen | |
| Blocked By: | Blocking: | ||
| Notify when closed: | Platform: | all | |
| Project: | ChimeraX |
Description
I am using the "sym" command in ChimeraX to expand an asymmetric unit to full icosahedral symmetry, MacOSX 10.14.4, ChimeraX latest daily build 0.9 (2019-05-01). The asymmetric unit is large, with 43 copies of the major capsid protein, but I have stripped it down to just 7656 alpha-carbon atoms total. It is read into ChimeraX from a PDB-format file. The sym command is: sym #1 i,2n5r center 559.3,559.3,559.3 I have quit other programs, and the Mac Pro has 64 GBytes of RAM and two AMD FirePro D700 graphics cards (6 GBytes each). The outcome varies from try to try: 1) Succeeds, and I can manipulate the large capsid model fine. 2) The screen returns to the login window, I login and ChimeraX is running with the correct result shown, and I can manipulate the large capsid model fine. 3) My session is logged out - I log in to a new session, have to start again. I haven't figured out what triggers each outcome, but I assume its a memory issue somewhere. A preferred outcome would be for ChimeraX to more gracefully indicate what the problem is, but I also wonder if using the PDB format is part of the problem, although expanding the symmetry like this should result in 459360 atoms, less that the 1,000,000 limit for PDB files. I assumed Chimera would perform poorly with this large model, but any advice appreciated. James Conway ---------------------------------------------------------------------- James Conway, PhD., Professor, Department of Structural Biology Director, Molecular Biophysics and Structural Biology Graduate Program University of Pittsburgh School of Medicine Biomedical Science Tower 3, Room 2047 3501 5th Ave Pittsburgh, PA 15213 U.S.A. Phone: +1-412-383-9847 Fax: +1-412-648-8998 Email: jxc100@pitt.edu Web: http://www.structbio.pitt.edu/index.php/12-faculty/18-james-conway ----------------------------------------------------------------------
Change History (3)
comment:1 by , 6 years ago
| Cc: | added |
|---|---|
| Component: | Unassigned → Higher-Order Structure |
| Owner: | set to |
| Platform: | → all |
| Project: | → ChimeraX |
| Status: | new → assigned |
comment:2 by , 6 years ago
| Resolution: | → nonchimerax |
|---|---|
| Status: | assigned → closed |
follow-up: 3 comment:3 by , 6 years ago
The solution is exactly right - simple lighting is quick and avoids the crashes.
I didn't know about the MacOSX graphics issue - good to know.
Thanks for the help. In the end, Chimera handled this ok - I thought it would be overwhelmed - and it also has the sym biometSet true switch that ChimeraX lacks. My goal here was to determine the symmetry used in the EM reconstruction (RELION's I4 == Chimera's i,n25r) and generate BIOMT matrices for depositing the asymmetric unit model in the PDB. All done!
Thanks again.
James Conway
----------------------------------------------------------------------
James Conway, PhD.,
Professor, Department of Structural Biology
Director, Molecular Biophysics and Structural Biology Graduate Program
University of Pittsburgh School of Medicine
Biomedical Science Tower 3, Room 2047
3501 5th Ave
Pittsburgh, PA 15213
U.S.A.
Phone: +1-412-383-9847
Fax: +1-412-648-8998
Email: jxc100@pitt.edu<mailto:jxc100@pitt.edu>
Web: http://www.structbio.pitt.edu/index.php/12-faculty/18-james-conway
----------------------------------------------------------------------
On 2 May, 2019, at 2:34 PM, ChimeraX <ChimeraX-bugs-admin@cgl.ucsf.edu<mailto:ChimeraX-bugs-admin@cgl.ucsf.edu>> wrote:
#1910: Crash with symmetry expansion
---------------------------------------------+---------------------
Reporter: James.Conway@… | Owner: goddard
Type: defect | Status: closed
Priority: normal | Milestone:
Component: Higher-Order Structure | Version:
Resolution: nonchimerax | Keywords:
Blocked By: | Blocking:
Notify when closed: | Platform: all
Project: ChimeraX |
---------------------------------------------+---------------------
Changes (by goddard):
* status: assigned => closed
* resolution: => nonchimerax
Comment:
This is a problem with macOS that I have seen before. If you ask the
graphics card to do too much work then it sometimes logs you out and or
kills ChimeraX. I believe the macOS rule is that if the graphics card is
not responding for 10 seconds then it assumes the graphics card crashed
and so it restarts the driver. This cause is reported in a macOS system
log file (I don't remember which one right now -- last I saw this was 6
months ago). I was not able to find any way to change the macOS behavior
-- it simply does a poor job handling heavy graphics computations. This
does not happen on Windows or Linux.
But there is a simple solution. The intense graphics computation is from
computing ambient occlusion lighting -- shadows cast from all directions.
The solution is to not use that lighting when you are trying to show this
virus capsid. Before you run the sym command use the command "light
simple", or use the equivalent icon in the toolbar (yellow light beam).
--
Ticket URL: <https://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fplato.cgl.ucsf.edu%2Ftrac%2FChimeraX%2Fticket%2F1910%23comment%3A2&data=02%7C01%7CJames.Conway%40pitt.edu%7Cfd481ed856ea45594dfc08d6cf2cd5d7%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1%7C1%7C636924188818847157&sdata=g40C8541UQcOMBDAW6dqTwlesG%2BZvHEmThnZWp3Mtog%3D&reserved=0>
ChimeraX <https://nam05.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.rbvi.ucsf.edu%2Fchimerax%2F&data=02%7C01%7CJames.Conway%40pitt.edu%7Cfd481ed856ea45594dfc08d6cf2cd5d7%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1%7C1%7C636924188818847157&sdata=Ysgelw9f12tse9sow%2Bs45%2BAIliHjfAa0cXMoLfQPLpU%3D&reserved=0>
ChimeraX Issue Tracker
Note:
See TracTickets
for help on using tickets.
This is a problem with macOS that I have seen before. If you ask the graphics card to do too much work then it sometimes logs you out and or kills ChimeraX. I believe the macOS rule is that if the graphics card is not responding for 10 seconds then it assumes the graphics card crashed and so it restarts the driver. This cause is reported in a macOS system log file (I don't remember which one right now -- last I saw this was 6 months ago). I was not able to find any way to change the macOS behavior -- it simply does a poor job handling heavy graphics computations. This does not happen on Windows or Linux.
But there is a simple solution. The intense graphics computation is from computing ambient occlusion lighting -- shadows cast from all directions. The solution is to not use that lighting when you are trying to show this virus capsid. Before you run the sym command use the command "light simple", or use the equivalent icon in the toolbar (yellow light beam).