Re: [chimerax-users] output mmCIF file
On Aug 31, 2021, at 1:50 PM, Liu, Xiangan <Xiangan.Liu@uth.tmc.edu> wrote:
Hi, I have issue to output the models I build in pdb formation to mmCIF formation. I did not provide the sequences in orginal pdb files. Could please let me know what is the problem and how to fix it? Thanks, Xiangan ---------- File "/usr/lib/ucsf-chimerax/lib/python3.8/site-packages/chimerax/mmcif/__init__.py", line 89, in save mmcif_write.write_mmcif(session, path, **kw) File "src/mmcif_write.pyx", line 145, in chimerax.mmcif.mmcif_write.write_mmcif File "src/mmcif_write.pyx", line 149, in chimerax.mmcif.mmcif_write.write_mmcif File "src/mmcif_write.pyx", line 542, in chimerax.mmcif.mmcif_write.save_structure KeyError: ('a1', 'RPGMMDSQEFS')
KeyError: ('a1', 'RPGMMDSQEFS')
File "src/mmcif_write.pyx", line 542, in chimerax.mmcif.mmcif_write.save_structure
Hi Xiangan, We recommend using the chimerax-users@cgl.ucsf.edu address for questions (CC'd on this message, so you don't need to do anything this time). The address you used is just for commercial licensing. Actually, if you get a traceback like this, you can usually use ChimeraX menu: Help... Report a Bug instead of sending e-mail. You may need to attach your session file so that we can reproduce the problem. I don't know the answer, so I'm using this address to share your question with the others. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Department of Pharmaceutical Chemistry University of California, San Francisco
Turns out that this bug is due to the chain ids in separate NMR models being explicitly changed to be different when they should all be same. That confuses the mmCIF writing code when it tries to write out the NMR ensemble, because it can't find the matching chain in different models. My guess is what was really wanted is a single model with each chain having a different chain id. That is easily done by using the combine command which takes a set of models and produces a new model and automatically names the chains. Then, when saving the mmCIF file, explicitly choose the new model. For example, if models #1 and #2 are the only open models and the one to combine, the command is: combine #1,2 which produces model #3. And the command to save that as a mmCIF file would be: save example.cif models #3 The backtrace is still a bug that needs to be fixed.... -- Greg On 8/31/2021 2:00 PM, Elaine Meng via ChimeraX-users wrote:
On Aug 31, 2021, at 1:50 PM, Liu, Xiangan <Xiangan.Liu@uth.tmc.edu> wrote:
Hi, I have issue to output the models I build in pdb formation to mmCIF formation. I did not provide the sequences in orginal pdb files. Could please let me know what is the problem and how to fix it? Thanks, Xiangan ---------- File "/usr/lib/ucsf-chimerax/lib/python3.8/site-packages/chimerax/mmcif/__init__.py", line 89, in save mmcif_write.write_mmcif(session, path, **kw) File "src/mmcif_write.pyx", line 145, in chimerax.mmcif.mmcif_write.write_mmcif File "src/mmcif_write.pyx", line 149, in chimerax.mmcif.mmcif_write.write_mmcif File "src/mmcif_write.pyx", line 542, in chimerax.mmcif.mmcif_write.save_structure KeyError: ('a1', 'RPGMMDSQEFS')
KeyError: ('a1', 'RPGMMDSQEFS')
File "src/mmcif_write.pyx", line 542, in chimerax.mmcif.mmcif_write.save_structure Hi Xiangan, We recommend using the chimerax-users@cgl.ucsf.edu address for questions (CC'd on this message, so you don't need to do anything this time). The address you used is just for commercial licensing.
Actually, if you get a traceback like this, you can usually use ChimeraX menu: Help... Report a Bug instead of sending e-mail. You may need to attach your session file so that we can reproduce the problem.
I don't know the answer, so I'm using this address to share your question with the others. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Department of Pharmaceutical Chemistry University of California, San Francisco
_______________________________________________ ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu Manage subscription: https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users
Hi Greg, It seems combine command does not work on my chimerax. I used rename #1,2 id #3, but when I save them into CIF file, it says "Not saving entity_poly_seq for non-authoritative sequences". I think my issue was probably caused by my changing the chain name using "setattr #2.3/b chain chain_id b12". The original name for all two 13-units are a or b. I want to name them as a1, b1,a2,b2,...etc. Any way to fix the issue? Thanks, Xiangan ________________________________ From: Greg Couch <gregc@cgl.ucsf.edu> Sent: Thursday, September 2, 2021 3:05 AM To: ChimeraX Users Help <ChimeraX-users@cgl.ucsf.edu> Cc: Liu, Xiangan <Xiangan.Liu@uth.tmc.edu> Subject: Re: [chimerax-users] output mmCIF file **** EXTERNAL EMAIL **** Turns out that this bug is due to the chain ids in separate NMR models being explicitly changed to be different when they should all be same. That confuses the mmCIF writing code when it tries to write out the NMR ensemble, because it can't find the matching chain in different models. My guess is what was really wanted is a single model with each chain having a different chain id. That is easily done by using the combine command which takes a set of models and produces a new model and automatically names the chains. Then, when saving the mmCIF file, explicitly choose the new model. For example, if models #1 and #2 are the only open models and the one to combine, the command is: combine #1,2 which produces model #3. And the command to save that as a mmCIF file would be: save example.cif models #3 The backtrace is still a bug that needs to be fixed.... -- Greg On 8/31/2021 2:00 PM, Elaine Meng via ChimeraX-users wrote:
On Aug 31, 2021, at 1:50 PM, Liu, Xiangan <Xiangan.Liu@uth.tmc.edu> wrote:
Hi, I have issue to output the models I build in pdb formation to mmCIF formation. I did not provide the sequences in orginal pdb files. Could please let me know what is the problem and how to fix it? Thanks, Xiangan ---------- File "/usr/lib/ucsf-chimerax/lib/python3.8/site-packages/chimerax/mmcif/https://urldefense.proofpoint.com/v2/url?u=http-3A__-5F-5Finit-5F-5F.py&d=Dw... ", line 89, in save mmcif_write.write_mmcif(session, path, **kw) File "src/mmcif_write.pyx", line 145, in chimerax.mmcif.mmcif_write.write_mmcif File "src/mmcif_write.pyx", line 149, in chimerax.mmcif.mmcif_write.write_mmcif File "src/mmcif_write.pyx", line 542, in chimerax.mmcif.mmcif_write.save_structure KeyError: ('a1', 'RPGMMDSQEFS')
KeyError: ('a1', 'RPGMMDSQEFS')
File "src/mmcif_write.pyx", line 542, in chimerax.mmcif.mmcif_write.save_structure Hi Xiangan, We recommend using the chimerax-users@cgl.ucsf.edu address for questions (CC'd on this message, so you don't need to do anything this time). The address you used is just for commercial licensing.
Actually, if you get a traceback like this, you can usually use ChimeraX menu: Help... Report a Bug instead of sending e-mail. You may need to attach your session file so that we can reproduce the problem.
I don't know the answer, so I'm using this address to share your question with the others. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Department of Pharmaceutical Chemistry University of California, San Francisco
_______________________________________________ ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu Manage subscription: https://urldefense.proofpoint.com/v2/url?u=https-3A__plato.cgl.ucsf.edu_mail...
Yes, the rename command does not do what you want. You need the combine command and that just in the daily build for now (sorry, I didn't realize that). So you'll need download and install the ChimeraX daily build. The combine command will be in the upcoming 1.3 release. The "Not saving entity_poly_seq for non-authoritative sequences" message is just a warning. ChimeraX benefits if the original data includes the full sequence information. Sequence searching is better, gaps can be depicted, etc. In your case, that information is missing. So the mmCIF writer does not put that information in the mmCIF file because it wasn't given in the first place. This is consistent with PHENIX's intermediate mmCIF files. If you want the mmCIF file to have that information anyway, use the "bestGuess true" option when saving the mmCIF file -- "save snafu.cif bestGuess true". -- Greg On 9/2/2021 9:46 AM, Liu, Xiangan wrote:
Hi Greg, It seems combine command does not work on my chimerax. I used rename #1,2 id #3, but when I save them into CIF file, it says "Not saving entity_poly_seq for non-authoritative sequences". I think my issue was probably caused by my changing the chain name using "setattr #2.3/b chain chain_id b12". The original name for all two 13-units are a or b. I want to name them as a1, b1,a2,b2,...etc. Any way to fix the issue? Thanks, Xiangan
------------------------------------------------------------------------ *From:* Greg Couch <gregc@cgl.ucsf.edu> *Sent:* Thursday, September 2, 2021 3:05 AM *To:* ChimeraX Users Help <ChimeraX-users@cgl.ucsf.edu> *Cc:* Liu, Xiangan <Xiangan.Liu@uth.tmc.edu> *Subject:* Re: [chimerax-users] output mmCIF file **** EXTERNAL EMAIL ****
Turns out that this bug is due to the chain ids in separate NMR models being explicitly changed to be different when they should all be same. That confuses the mmCIF writing code when it tries to write out the NMR ensemble, because it can't find the matching chain in different models.
My guess is what was really wanted is a single model with each chain having a different chain id. That is easily done by using the combine command which takes a set of models and produces a new model and automatically names the chains. Then, when saving the mmCIF file, explicitly choose the new model. For example, if models #1 and #2 are the only open models and the one to combine, the command is:
combine #1,2
which produces model #3. And the command to save that as a mmCIF file would be:
save example.cif models #3
The backtrace is still a bug that needs to be fixed....
-- Greg
On 8/31/2021 2:00 PM, Elaine Meng via ChimeraX-users wrote:
On Aug 31, 2021, at 1:50 PM, Liu, Xiangan <Xiangan.Liu@uth.tmc.edu> wrote:
Hi, I have issue to output the models I build in pdb formation to mmCIF formation. I did not provide the sequences in orginal pdb files. Could please let me know what is the problem and how to fix it? Thanks, Xiangan ---------- File "/usr/lib/ucsf-chimerax/lib/python3.8/site-packages/chimerax/mmcif/https://urldefense.proofpoint.com/v2/url?u=http-3A__-5F-5Finit-5F-5F.py&d=Dw... ", line 89, in save mmcif_write.write_mmcif(session, path, **kw) File "src/mmcif_write.pyx", line 145, in chimerax.mmcif.mmcif_write.write_mmcif File "src/mmcif_write.pyx", line 149, in chimerax.mmcif.mmcif_write.write_mmcif File "src/mmcif_write.pyx", line 542, in chimerax.mmcif.mmcif_write.save_structure KeyError: ('a1', 'RPGMMDSQEFS')
KeyError: ('a1', 'RPGMMDSQEFS')
File "src/mmcif_write.pyx", line 542, in chimerax.mmcif.mmcif_write.save_structure Hi Xiangan, We recommend using the chimerax-users@cgl.ucsf.edu address for questions (CC'd on this message, so you don't need to do anything this time). The address you used is just for commercial licensing.
Actually, if you get a traceback like this, you can usually use ChimeraX menu: Help... Report a Bug instead of sending e-mail. You may need to attach your session file so that we can reproduce the problem.
I don't know the answer, so I'm using this address to share your question with the others. Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Department of Pharmaceutical Chemistry University of California, San Francisco
_______________________________________________ ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu Manage subscription:
https://urldefense.proofpoint.com/v2/url?u=https-3A__plato.cgl.ucsf.edu_mail... <https://urldefense.proofpoint.com/v2/url?u=https-3A__plato.cgl.ucsf.edu_mail...>
participants (3)
-
Elaine Meng -
Greg Couch -
Liu, Xiangan