[chimera-dev] Update to DelphiViewer Chimera plugin

Wed Jan 8 11:48:49 PST 2020

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 at 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 at 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 at 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.phi
>>
>> 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 at 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 at cgl.ucsf.edu
>>> http://www.rbvi.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 at cgl.ucsf.edu
> http://www.rbvi.ucsf.edu/mailman/listinfo/chimera-dev
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-dev/attachments/20200108/513fccd0/attachment.html>