Re: [chimera-dev] Update to DelphiViewer Chimera plugin
Hi Emil,
We will update Chimera to use CUBE format with DelPhi before our next release.
The link I sent you to the Chimera bug report gives the full DelPhi output and DelPhi version that produced the incorrect file.
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169 http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
I've attached it here as a text file. I asked the user to contact your lab a several days ago but I don't know if they have time.
Thanks!
Tom
On Jan 8, 2020, at 9:56 AM, Emil Alexov ealexov@g.clemson.edu wrote:
Hi Tom,
Please switch to CUBE format. Regarding the user, who knows how the file was generated. Let them contact us and we will figure out what was wrong. In any case, I agree with you that wrong format file should not be generated. Ask the user to contact us and we will let you know the outcome.
Emil
On Wed, Jan 8, 2020 at 12:33 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Emil,
I agree it is probably a misunderstanding. The CUBE format is fine and I can switch new Chimera versions to use it. The trouble is that a user sent me a *.phi file written by DelPhi 7 that does not follow the format specification for .phi.
http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phihttp://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.p... http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phi
as part of Chimera bug report
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169 http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
This is a binary file so it is not an ascii CUBE file. Looking at the strings at the beginning of the file it appears to be .phi format without required record lengths. If this change to the DelPhi .phi file output was not intended then either it should be fixed to output a correct .phi format, or DelPhi should be made to not output .phi at all (just give an error that it is no longer supported). As it is now, DelPhi outputs an incorrect file.
Tom
On Jan 8, 2020, at 3:48 AM, Emil Alexov wrote:
Tom,
This seems to me is misunderstanding. We purposely decided to make the new delphi version to write PHI map file into CUBE format, because this is the format that all existing visualization packages take. This is documented in DelPhi manual and nowadays our users indeed prefer CUBE format versus the old style binary file.
However, if this is too much trouble for Chimera maintenance and development, we will make an effort to include binary option in the new delphi. Let's us know what are your thoughts about this.
Emil Alexov
On Tue, Jan 7, 2020 at 6:38 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Shailesh,
Thanks. I see your changes to the Chimera DelphiViewer plugin make it write out a Gaussian cube file instead instead of a .phi file. While that helps a little it would be much better if new DelPhi wrote out a .phi file following the online documentation for that file format. This should be trivial to do even from C++ (while the original format used Fortran records, they are just a 4 byte count for each record, followed by the binary data). Having DelPhi write out a different incompatible .phi file as it dose not breaks all software that used that format. We can't take the tens of thousands of copies of Chimera out of circulation that read .phi files. Also even with your change to make Chimera run DelPhi outputting a cube file, users will run DelPhi outside of Chimera, then try to open the .phi file and get an error.
If the .phi format is just going to remain broken with new DelPhi versions writing an incompatible format I think we just need to remove it from Chimera and our new program ChimeraX. If your new .phi format is in use by new software and cannot be fixed I'd suggest that you use a different file suffix so software can correctly read the new or old format.
Tom
On Jan 7, 2020, at 10:55 AM, Shailesh Kumar Panday wrote:
Hello Chimera Developers,
We have made some changes in the DelphiViewer plugin to enable reading and writing Delphi potential in "Gaussian cube grid" format. The older binary "Delphi or GRASP potential" format files are no longer supported in Delphi. Hence plugin was unable to parse the potential file, the attached version fixes this issue.
Thank you,
Shailesh Kumar Panday Postdoctoral Fellow Computational Biophysics and Bioinformatics Lab Department of Physics and As Clemson University, SC 29634 Lab: http://compbio.clemson.edu http://compbio.clemson.edu/ <DelphiViewer.tar>_______________________________________________ Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu mailto:Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu http://compbio.clemson.edu/
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu http://compbio.clemson.edu/
Hi Emil,
A few more thoughts on Chimera display of DelPhi electrostatics. Chimera has for a decade been able to read CUBE format and Elaine Meng here checked that DelPhi CUBE format already opens correctly in Chimera. But the Chimera electrostatic coloring tool does not automatically appear when opening a CUBE file because that format is primarily used for molecular orbitals (output by Gaussian I think). So a user will have to know how to find the electrostatic coloring tool in Chimera to use the opened CUBE file. Another small drawback of the ascii CUBE file versus the older binary .phi file is that it is slower to read (maybe about 5x slower, or 10 seconds instead of 2 seconds for larger molecules).
Regarding changing Chimera to run DelPhi and produce CUBE format output -- there is the problem that Chimera may be running an older DelPhi that does not write CUBE format. It uses whatever DelPhi the user installed on their computer. When was CUBE output added to DelPhi? Changing Chimera to request CUBE output would break Chimera DelPhi runs with older DelPhi.
To mitigate these problems it would be helpful if DelPhi outputted the correct .phi binary format. The format is extremely simple, the Chimera code that reads it is less than 2 pages.
Tom
On Jan 8, 2020, at 10:46 AM, Tom Goddard wrote:
Hi Emil,
We will update Chimera to use CUBE format with DelPhi before our next release.
The link I sent you to the Chimera bug report gives the full DelPhi output and DelPhi version that produced the incorrect file.
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169 http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
I've attached it here as a text file. I asked the user to contact your lab a several days ago but I don't know if they have time.
Thanks!
Tom
<delphi.txt>
On Jan 8, 2020, at 9:56 AM, Emil Alexov <ealexov@g.clemson.edu mailto:ealexov@g.clemson.edu> wrote:
Hi Tom,
Please switch to CUBE format. Regarding the user, who knows how the file was generated. Let them contact us and we will figure out what was wrong. In any case, I agree with you that wrong format file should not be generated. Ask the user to contact us and we will let you know the outcome.
Emil
On Wed, Jan 8, 2020 at 12:33 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Emil,
I agree it is probably a misunderstanding. The CUBE format is fine and I can switch new Chimera versions to use it. The trouble is that a user sent me a *.phi file written by DelPhi 7 that does not follow the format specification for .phi.
http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phihttp://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.p... http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phi
as part of Chimera bug report
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169 http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
This is a binary file so it is not an ascii CUBE file. Looking at the strings at the beginning of the file it appears to be .phi format without required record lengths. If this change to the DelPhi .phi file output was not intended then either it should be fixed to output a correct .phi format, or DelPhi should be made to not output .phi at all (just give an error that it is no longer supported). As it is now, DelPhi outputs an incorrect file.
Tom
On Jan 8, 2020, at 3:48 AM, Emil Alexov wrote:
Tom,
This seems to me is misunderstanding. We purposely decided to make the new delphi version to write PHI map file into CUBE format, because this is the format that all existing visualization packages take. This is documented in DelPhi manual and nowadays our users indeed prefer CUBE format versus the old style binary file.
However, if this is too much trouble for Chimera maintenance and development, we will make an effort to include binary option in the new delphi. Let's us know what are your thoughts about this.
Emil Alexov
On Tue, Jan 7, 2020 at 6:38 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Shailesh,
Thanks. I see your changes to the Chimera DelphiViewer plugin make it write out a Gaussian cube file instead instead of a .phi file. While that helps a little it would be much better if new DelPhi wrote out a .phi file following the online documentation for that file format. This should be trivial to do even from C++ (while the original format used Fortran records, they are just a 4 byte count for each record, followed by the binary data). Having DelPhi write out a different incompatible .phi file as it dose not breaks all software that used that format. We can't take the tens of thousands of copies of Chimera out of circulation that read .phi files. Also even with your change to make Chimera run DelPhi outputting a cube file, users will run DelPhi outside of Chimera, then try to open the .phi file and get an error.
If the .phi format is just going to remain broken with new DelPhi versions writing an incompatible format I think we just need to remove it from Chimera and our new program ChimeraX. If your new .phi format is in use by new software and cannot be fixed I'd suggest that you use a different file suffix so software can correctly read the new or old format.
Tom
On Jan 7, 2020, at 10:55 AM, Shailesh Kumar Panday wrote:
Hello Chimera Developers,
We have made some changes in the DelphiViewer plugin to enable reading and writing Delphi potential in "Gaussian cube grid" format. The older binary "Delphi or GRASP potential" format files are no longer supported in Delphi. Hence plugin was unable to parse the potential file, the attached version fixes this issue.
Thank you,
Shailesh Kumar Panday Postdoctoral Fellow Computational Biophysics and Bioinformatics Lab Department of Physics and As Clemson University, SC 29634 Lab: http://compbio.clemson.edu http://compbio.clemson.edu/ <DelphiViewer.tar>_______________________________________________ Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu mailto:Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu http://compbio.clemson.edu/
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu http://compbio.clemson.edu/
Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
Here is Elaine's description of how to do electrostatic coloring in Chimera with a CUBE file.
http://plato.cgl.ucsf.edu/pipermail/chimera-users/2019-December/016397.html http://plato.cgl.ucsf.edu/pipermail/chimera-users/2019-December/016397.html
Tom
On Jan 8, 2020, at 11:38 AM, Tom Goddard goddard@sonic.net wrote:
Hi Emil,
A few more thoughts on Chimera display of DelPhi electrostatics. Chimera has for a decade been able to read CUBE format and Elaine Meng here checked that DelPhi CUBE format already opens correctly in Chimera. But the Chimera electrostatic coloring tool does not automatically appear when opening a CUBE file because that format is primarily used for molecular orbitals (output by Gaussian I think). So a user will have to know how to find the electrostatic coloring tool in Chimera to use the opened CUBE file. Another small drawback of the ascii CUBE file versus the older binary .phi file is that it is slower to read (maybe about 5x slower, or 10 seconds instead of 2 seconds for larger molecules).
Regarding changing Chimera to run DelPhi and produce CUBE format output -- there is the problem that Chimera may be running an older DelPhi that does not write CUBE format. It uses whatever DelPhi the user installed on their computer. When was CUBE output added to DelPhi? Changing Chimera to request CUBE output would break Chimera DelPhi runs with older DelPhi.
To mitigate these problems it would be helpful if DelPhi outputted the correct .phi binary format. The format is extremely simple, the Chimera code that reads it is less than 2 pages.
Tom
On Jan 8, 2020, at 10:46 AM, Tom Goddard wrote:
Hi Emil,
We will update Chimera to use CUBE format with DelPhi before our next release.
The link I sent you to the Chimera bug report gives the full DelPhi output and DelPhi version that produced the incorrect file.
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169 http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
I've attached it here as a text file. I asked the user to contact your lab a several days ago but I don't know if they have time.
Thanks!
Tom
<delphi.txt>
On Jan 8, 2020, at 9:56 AM, Emil Alexov <ealexov@g.clemson.edu mailto:ealexov@g.clemson.edu> wrote:
Hi Tom,
Please switch to CUBE format. Regarding the user, who knows how the file was generated. Let them contact us and we will figure out what was wrong. In any case, I agree with you that wrong format file should not be generated. Ask the user to contact us and we will let you know the outcome.
Emil
On Wed, Jan 8, 2020 at 12:33 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Emil,
I agree it is probably a misunderstanding. The CUBE format is fine and I can switch new Chimera versions to use it. The trouble is that a user sent me a *.phi file written by DelPhi 7 that does not follow the format specification for .phi.
http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phihttp://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.p... http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phi
as part of Chimera bug report
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169 http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
This is a binary file so it is not an ascii CUBE file. Looking at the strings at the beginning of the file it appears to be .phi format without required record lengths. If this change to the DelPhi .phi file output was not intended then either it should be fixed to output a correct .phi format, or DelPhi should be made to not output .phi at all (just give an error that it is no longer supported). As it is now, DelPhi outputs an incorrect file.
Tom
On Jan 8, 2020, at 3:48 AM, Emil Alexov wrote:
Tom,
This seems to me is misunderstanding. We purposely decided to make the new delphi version to write PHI map file into CUBE format, because this is the format that all existing visualization packages take. This is documented in DelPhi manual and nowadays our users indeed prefer CUBE format versus the old style binary file.
However, if this is too much trouble for Chimera maintenance and development, we will make an effort to include binary option in the new delphi. Let's us know what are your thoughts about this.
Emil Alexov
On Tue, Jan 7, 2020 at 6:38 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Shailesh,
Thanks. I see your changes to the Chimera DelphiViewer plugin make it write out a Gaussian cube file instead instead of a .phi file. While that helps a little it would be much better if new DelPhi wrote out a .phi file following the online documentation for that file format. This should be trivial to do even from C++ (while the original format used Fortran records, they are just a 4 byte count for each record, followed by the binary data). Having DelPhi write out a different incompatible .phi file as it dose not breaks all software that used that format. We can't take the tens of thousands of copies of Chimera out of circulation that read .phi files. Also even with your change to make Chimera run DelPhi outputting a cube file, users will run DelPhi outside of Chimera, then try to open the .phi file and get an error.
If the .phi format is just going to remain broken with new DelPhi versions writing an incompatible format I think we just need to remove it from Chimera and our new program ChimeraX. If your new .phi format is in use by new software and cannot be fixed I'd suggest that you use a different file suffix so software can correctly read the new or old format.
Tom
On Jan 7, 2020, at 10:55 AM, Shailesh Kumar Panday wrote:
Hello Chimera Developers,
We have made some changes in the DelphiViewer plugin to enable reading and writing Delphi potential in "Gaussian cube grid" format. The older binary "Delphi or GRASP potential" format files are no longer supported in Delphi. Hence plugin was unable to parse the potential file, the attached version fixes this issue.
Thank you,
Shailesh Kumar Panday Postdoctoral Fellow Computational Biophysics and Bioinformatics Lab Department of Physics and As Clemson University, SC 29634 Lab: http://compbio.clemson.edu http://compbio.clemson.edu/ <DelphiViewer.tar>_______________________________________________ Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu mailto:Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu http://compbio.clemson.edu/
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu http://compbio.clemson.edu/
Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu mailto:Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
Hi Tom,
Good points. We will enable the binary format in a day or so, and let you know.
Emil
On Wed, Jan 8, 2020, 2:38 PM Tom Goddard goddard@sonic.net wrote:
Hi Emil,
A few more thoughts on Chimera display of DelPhi electrostatics. Chimera has for a decade been able to read CUBE format and Elaine Meng here checked that DelPhi CUBE format already opens correctly in Chimera. But the Chimera electrostatic coloring tool does not automatically appear when opening a CUBE file because that format is primarily used for molecular orbitals (output by Gaussian I think). So a user will have to know how to find the electrostatic coloring tool in Chimera to use the opened CUBE file. Another small drawback of the ascii CUBE file versus the older binary .phi file is that it is slower to read (maybe about 5x slower, or 10 seconds instead of 2 seconds for larger molecules).
Regarding changing Chimera to run DelPhi and produce CUBE format output -- there is the problem that Chimera may be running an older DelPhi that does not write CUBE format. It uses whatever DelPhi the user installed on their computer. When was CUBE output added to DelPhi? Changing Chimera to request CUBE output would break Chimera DelPhi runs with older DelPhi.
To mitigate these problems it would be helpful if DelPhi outputted the correct .phi binary format. The format is extremely simple, the Chimera code that reads it is less than 2 pages.
Tom
On Jan 8, 2020, at 10:46 AM, Tom Goddard wrote:
Hi Emil,
We will update Chimera to use CUBE format with DelPhi before our next release.
The link I sent you to the Chimera bug report gives the full DelPhi output and DelPhi version that produced the incorrect file.
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
I've attached it here as a text file. I asked the user to contact your lab a several days ago but I don't know if they have time.
Thanks!
Tom
<delphi.txt>
On Jan 8, 2020, at 9:56 AM, Emil Alexov ealexov@g.clemson.edu wrote:
Hi Tom,
Please switch to CUBE format. Regarding the user, who knows how the file was generated. Let them contact us and we will figure out what was wrong. In any case, I agree with you that wrong format file should not be generated. Ask the user to contact us and we will let you know the outcome.
Emil
On Wed, Jan 8, 2020 at 12:33 PM Tom Goddard goddard@sonic.net wrote:
Hi Emil,
I agree it is probably a misunderstanding. The CUBE format is fine and I can switch new Chimera versions to use it. The trouble is that a user sent me a *.phi file written by DelPhi 7 that does not follow the format specification for .phi.
http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phi http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.p...
as part of Chimera bug report
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
This is a binary file so it is not an ascii CUBE file. Looking at the strings at the beginning of the file it appears to be .phi format without required record lengths. If this change to the DelPhi .phi file output was not intended then either it should be fixed to output a correct .phi format, or DelPhi should be made to not output .phi at all (just give an error that it is no longer supported). As it is now, DelPhi outputs an incorrect file.
Tom
On Jan 8, 2020, at 3:48 AM, Emil Alexov wrote:
Tom,
This seems to me is misunderstanding. We purposely decided to make the new delphi version to write PHI map file into CUBE format, because this is the format that all existing visualization packages take. This is documented in DelPhi manual and nowadays our users indeed prefer CUBE format versus the old style binary file.
However, if this is too much trouble for Chimera maintenance and development, we will make an effort to include binary option in the new delphi. Let's us know what are your thoughts about this.
Emil Alexov
On Tue, Jan 7, 2020 at 6:38 PM Tom Goddard goddard@sonic.net wrote:
Hi Shailesh,
Thanks. I see your changes to the Chimera DelphiViewer plugin make it write out a Gaussian cube file instead instead of a .phi file. While that helps a little it would be much better if new DelPhi wrote out a .phi file following the online documentation for that file format. This should be trivial to do even from C++ (while the original format used Fortran records, they are just a 4 byte count for each record, followed by the binary data). Having DelPhi write out a different incompatible .phi file as it dose not breaks all software that used that format. We can't take the tens of thousands of copies of Chimera out of circulation that read .phi files. Also even with your change to make Chimera run DelPhi outputting a cube file, users will run DelPhi outside of Chimera, then try to open the .phi file and get an error.
If the .phi format is just going to remain broken with new DelPhi versions writing an incompatible format I think we just need to remove it from Chimera and our new program ChimeraX. If your new .phi format is in use by new software and cannot be fixed I'd suggest that you use a different file suffix so software can correctly read the new or old format.
Tom
On Jan 7, 2020, at 10:55 AM, Shailesh Kumar Panday wrote:
Hello Chimera Developers,
We have made some changes in the DelphiViewer plugin to enable reading and writing Delphi potential in "Gaussian cube grid" format. The older binary "Delphi or GRASP potential" format files are no longer supported in Delphi. Hence plugin was unable to parse the potential file, the attached version fixes this issue.
Thank you,
Shailesh Kumar Panday Postdoctoral Fellow Computational Biophysics and Bioinformatics Lab Department of Physics and As Clemson University, SC 29634 Lab: http://compbio.clemson.edu
<DelphiViewer.tar>_______________________________________________ Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu
Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
Hi Tom,
The issue was fixed and now the new Delphi release generates phi map file in desired format. We provide executable for Linux, Mac and Windows.
If you receive Chimera user inquire about phi map, tell them to download the newest Delphi.
Emil
On Wed, Jan 8, 2020, 2:38 PM Tom Goddard goddard@sonic.net wrote:
Hi Emil,
A few more thoughts on Chimera display of DelPhi electrostatics. Chimera has for a decade been able to read CUBE format and Elaine Meng here checked that DelPhi CUBE format already opens correctly in Chimera. But the Chimera electrostatic coloring tool does not automatically appear when opening a CUBE file because that format is primarily used for molecular orbitals (output by Gaussian I think). So a user will have to know how to find the electrostatic coloring tool in Chimera to use the opened CUBE file. Another small drawback of the ascii CUBE file versus the older binary .phi file is that it is slower to read (maybe about 5x slower, or 10 seconds instead of 2 seconds for larger molecules).
Regarding changing Chimera to run DelPhi and produce CUBE format output -- there is the problem that Chimera may be running an older DelPhi that does not write CUBE format. It uses whatever DelPhi the user installed on their computer. When was CUBE output added to DelPhi? Changing Chimera to request CUBE output would break Chimera DelPhi runs with older DelPhi.
To mitigate these problems it would be helpful if DelPhi outputted the correct .phi binary format. The format is extremely simple, the Chimera code that reads it is less than 2 pages.
Tom
On Jan 8, 2020, at 10:46 AM, Tom Goddard wrote:
Hi Emil,
We will update Chimera to use CUBE format with DelPhi before our next release.
The link I sent you to the Chimera bug report gives the full DelPhi output and DelPhi version that produced the incorrect file.
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
I've attached it here as a text file. I asked the user to contact your lab a several days ago but I don't know if they have time.
Thanks!
Tom
<delphi.txt>
On Jan 8, 2020, at 9:56 AM, Emil Alexov ealexov@g.clemson.edu wrote:
Hi Tom,
Please switch to CUBE format. Regarding the user, who knows how the file was generated. Let them contact us and we will figure out what was wrong. In any case, I agree with you that wrong format file should not be generated. Ask the user to contact us and we will let you know the outcome.
Emil
On Wed, Jan 8, 2020 at 12:33 PM Tom Goddard goddard@sonic.net wrote:
Hi Emil,
I agree it is probably a misunderstanding. The CUBE format is fine and I can switch new Chimera versions to use it. The trouble is that a user sent me a *.phi file written by DelPhi 7 that does not follow the format specification for .phi.
http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phi http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.p...
as part of Chimera bug report
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
This is a binary file so it is not an ascii CUBE file. Looking at the strings at the beginning of the file it appears to be .phi format without required record lengths. If this change to the DelPhi .phi file output was not intended then either it should be fixed to output a correct .phi format, or DelPhi should be made to not output .phi at all (just give an error that it is no longer supported). As it is now, DelPhi outputs an incorrect file.
Tom
On Jan 8, 2020, at 3:48 AM, Emil Alexov wrote:
Tom,
This seems to me is misunderstanding. We purposely decided to make the new delphi version to write PHI map file into CUBE format, because this is the format that all existing visualization packages take. This is documented in DelPhi manual and nowadays our users indeed prefer CUBE format versus the old style binary file.
However, if this is too much trouble for Chimera maintenance and development, we will make an effort to include binary option in the new delphi. Let's us know what are your thoughts about this.
Emil Alexov
On Tue, Jan 7, 2020 at 6:38 PM Tom Goddard goddard@sonic.net wrote:
Hi Shailesh,
Thanks. I see your changes to the Chimera DelphiViewer plugin make it write out a Gaussian cube file instead instead of a .phi file. While that helps a little it would be much better if new DelPhi wrote out a .phi file following the online documentation for that file format. This should be trivial to do even from C++ (while the original format used Fortran records, they are just a 4 byte count for each record, followed by the binary data). Having DelPhi write out a different incompatible .phi file as it dose not breaks all software that used that format. We can't take the tens of thousands of copies of Chimera out of circulation that read .phi files. Also even with your change to make Chimera run DelPhi outputting a cube file, users will run DelPhi outside of Chimera, then try to open the .phi file and get an error.
If the .phi format is just going to remain broken with new DelPhi versions writing an incompatible format I think we just need to remove it from Chimera and our new program ChimeraX. If your new .phi format is in use by new software and cannot be fixed I'd suggest that you use a different file suffix so software can correctly read the new or old format.
Tom
On Jan 7, 2020, at 10:55 AM, Shailesh Kumar Panday wrote:
Hello Chimera Developers,
We have made some changes in the DelphiViewer plugin to enable reading and writing Delphi potential in "Gaussian cube grid" format. The older binary "Delphi or GRASP potential" format files are no longer supported in Delphi. Hence plugin was unable to parse the potential file, the attached version fixes this issue.
Thank you,
Shailesh Kumar Panday Postdoctoral Fellow Computational Biophysics and Bioinformatics Lab Department of Physics and As Clemson University, SC 29634 Lab: http://compbio.clemson.edu
<DelphiViewer.tar>_______________________________________________ Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu
Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
Great! Thanks!
Tom
On Jan 17, 2020, at 11:13 AM, Emil Alexov ealexov@g.clemson.edu wrote:
Hi Tom,
The issue was fixed and now the new Delphi release generates phi map file in desired format. We provide executable for Linux, Mac and Windows.
If you receive Chimera user inquire about phi map, tell them to download the newest Delphi.
Emil
On Wed, Jan 8, 2020, 2:38 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Emil,
A few more thoughts on Chimera display of DelPhi electrostatics. Chimera has for a decade been able to read CUBE format and Elaine Meng here checked that DelPhi CUBE format already opens correctly in Chimera. But the Chimera electrostatic coloring tool does not automatically appear when opening a CUBE file because that format is primarily used for molecular orbitals (output by Gaussian I think). So a user will have to know how to find the electrostatic coloring tool in Chimera to use the opened CUBE file. Another small drawback of the ascii CUBE file versus the older binary .phi file is that it is slower to read (maybe about 5x slower, or 10 seconds instead of 2 seconds for larger molecules).
Regarding changing Chimera to run DelPhi and produce CUBE format output -- there is the problem that Chimera may be running an older DelPhi that does not write CUBE format. It uses whatever DelPhi the user installed on their computer. When was CUBE output added to DelPhi? Changing Chimera to request CUBE output would break Chimera DelPhi runs with older DelPhi.
To mitigate these problems it would be helpful if DelPhi outputted the correct .phi binary format. The format is extremely simple, the Chimera code that reads it is less than 2 pages.
Tom
On Jan 8, 2020, at 10:46 AM, Tom Goddard wrote:
Hi Emil,
We will update Chimera to use CUBE format with DelPhi before our next release.
The link I sent you to the Chimera bug report gives the full DelPhi output and DelPhi version that produced the incorrect file.
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169 http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
I've attached it here as a text file. I asked the user to contact your lab a several days ago but I don't know if they have time.
Thanks!
Tom
<delphi.txt>
On Jan 8, 2020, at 9:56 AM, Emil Alexov <ealexov@g.clemson.edu mailto:ealexov@g.clemson.edu> wrote:
Hi Tom,
Please switch to CUBE format. Regarding the user, who knows how the file was generated. Let them contact us and we will figure out what was wrong. In any case, I agree with you that wrong format file should not be generated. Ask the user to contact us and we will let you know the outcome.
Emil
On Wed, Jan 8, 2020 at 12:33 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Emil,
I agree it is probably a misunderstanding. The CUBE format is fine and I can switch new Chimera versions to use it. The trouble is that a user sent me a *.phi file written by DelPhi 7 that does not follow the format specification for .phi.
http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phihttp://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.p... http://plato.cgl.ucsf.edu/trac/chimera/raw-attachment/ticket/17169/default.phi
as part of Chimera bug report
http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169 http://plato.cgl.ucsf.edu/trac/chimera/ticket/17169
This is a binary file so it is not an ascii CUBE file. Looking at the strings at the beginning of the file it appears to be .phi format without required record lengths. If this change to the DelPhi .phi file output was not intended then either it should be fixed to output a correct .phi format, or DelPhi should be made to not output .phi at all (just give an error that it is no longer supported). As it is now, DelPhi outputs an incorrect file.
Tom
On Jan 8, 2020, at 3:48 AM, Emil Alexov wrote:
Tom,
This seems to me is misunderstanding. We purposely decided to make the new delphi version to write PHI map file into CUBE format, because this is the format that all existing visualization packages take. This is documented in DelPhi manual and nowadays our users indeed prefer CUBE format versus the old style binary file.
However, if this is too much trouble for Chimera maintenance and development, we will make an effort to include binary option in the new delphi. Let's us know what are your thoughts about this.
Emil Alexov
On Tue, Jan 7, 2020 at 6:38 PM Tom Goddard <goddard@sonic.net mailto:goddard@sonic.net> wrote: Hi Shailesh,
Thanks. I see your changes to the Chimera DelphiViewer plugin make it write out a Gaussian cube file instead instead of a .phi file. While that helps a little it would be much better if new DelPhi wrote out a .phi file following the online documentation for that file format. This should be trivial to do even from C++ (while the original format used Fortran records, they are just a 4 byte count for each record, followed by the binary data). Having DelPhi write out a different incompatible .phi file as it dose not breaks all software that used that format. We can't take the tens of thousands of copies of Chimera out of circulation that read .phi files. Also even with your change to make Chimera run DelPhi outputting a cube file, users will run DelPhi outside of Chimera, then try to open the .phi file and get an error.
If the .phi format is just going to remain broken with new DelPhi versions writing an incompatible format I think we just need to remove it from Chimera and our new program ChimeraX. If your new .phi format is in use by new software and cannot be fixed I'd suggest that you use a different file suffix so software can correctly read the new or old format.
Tom
On Jan 7, 2020, at 10:55 AM, Shailesh Kumar Panday wrote:
Hello Chimera Developers,
We have made some changes in the DelphiViewer plugin to enable reading and writing Delphi potential in "Gaussian cube grid" format. The older binary "Delphi or GRASP potential" format files are no longer supported in Delphi. Hence plugin was unable to parse the potential file, the attached version fixes this issue.
Thank you,
Shailesh Kumar Panday Postdoctoral Fellow Computational Biophysics and Bioinformatics Lab Department of Physics and As Clemson University, SC 29634 Lab: http://compbio.clemson.edu http://compbio.clemson.edu/ <DelphiViewer.tar>_______________________________________________ Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu mailto:Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu http://compbio.clemson.edu/
-- Emil Alexov Professor, Department of Physics Professor, Department of Material Sciences Professor, School of Nursing: Healthcare Genetics Faculty Scholar at Clemson School of Health Research Clemson University Lab: http://compbio.clemson.edu http://compbio.clemson.edu/
Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu mailto:Chimera-dev@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-dev
participants (2)
-
Emil Alexov -
Tom Goddard