Re: [PyMOL] Pymol and APBS

[Shiven S. <shi...@gm...>]

Dear Michael,

First, let me thank you for your time and effort in providing the community
with your contributions. I do sincerely appreciate your hard work and your
willingness to share it with us all. Without the contributions of users such
as yourself, PyMOL's stature and utility would be far less than what it is
now.

I have used the APBS plugin in the past, successfully and without problems.
And yes, it has been helpful in solving the problem(s) at hand.

As for the "knocking down", it was more of a comparison with VASCo than a
direct criticism of your work. I am no expert at programming PyMOL/python
plugins but I have used quite a few over the years. All I speak is from my
own "user experience" when I say that some plugins work better than others.
VASCo, in my experience, happens to belong to the "better" category.


Now why the "better plugins" work the way do may depend on many factors but
some that I feel important are listed below (nothing specifically
for/against the APBS plugin).

I only speculate when I write the following, but in hopes that it may be
considered "constructive feedback":

(1). Whenever possible, farm out the heavy lifting to external executables
(apbs, python scripts, black magic shamans etc.)...

(2). Clearly ask for all the defaults a user must specify at "first" run
(executable location, temporary directory, other knick-knacks your plugin
requires).... Do not allow execution to begin unless you are reasonably sure
you have everything needed to properly finish the job.

(3). And, for heaven's sake, save those user-specified defaults for the rest
of posterity in a location (pymolrc, yourpluginrc) that your plugin can
always find... never, ever make a user specify such mundane things over and
over again... nothing kills the "user experience" quicker than that.

(4). Overwrite result files only if you have to... create a new directory
for every new run (and clearly display the location)...
Remember, (storage) space is cheap and (a user's) time is expensive, respect
that.

(5). If possible, give the users a clear indication of the status of the
back-end job doing the heavy lifting...

(6). Once you are done running the back-end job(s), prompt the user
immediately and (if agreed) automatically load the results.

(7). If your back-end job fails, parse the error in your plugin and inform
the user where in the entire process the error occurred, why it may have
occurred and what the user should do to prevent this error from occurring
again. If it is an error your plugin does not understand let the user know,
clearly and loudly.  If an error log file, hopefully text, was written,
specify its location (full path) in your error prompts. Prefrably, display
that error file in the system %editor% so that the user does not have to go
hunting for it.

...and I'll leave it at that...  ;-)

I welcome your feedback and views of other community members on this.

Thank you!

With best regards,
Shiven

P.S.  Apologies again if my criticism hit a personal note. Like everyone
else writing to the list and answering questions, my aim was to help a
fellow user "get that job done".


---------------


On Thu, Apr 15, 2010 at 12:00 PM, Michael Lerner <mgl...@gm...> wrote:

>
>
> On Thu, Apr 15, 2010 at 10:30 AM, Shiven Shandilya <
> shi...@gm...> wrote:
>
>> Sorry to be knocking down APBS+PyMOL integration like this, but IMHO VASCo
>> does a stellar job as far as integrating outside programs with PyMOL goes.
>>
>>
> I initially wrote the PyMOL/APBS plugin for my own use and I now
> maintain/develop it as a community resource in my limited free time. Despite
> this, several hundred users (at least) have found it quite helpful in their
> research, and I've included many newer features based on their feedback. The
> plugin is certainly not perfect, and I don't mind people "knocking it down,"
> but I appreciate it when that knocking down comes with constructive
> feedback.
>
> Additionally, unlike VASCo, the PyMOL/APBS/PDB2PQR toolchain is provided as
> completely free and open source suite of programs. For many reasons, I
> believe that free and open source software is vital to to good science. I
> would strongly urge users to consider such issues before settling on a suite
> of software products, especially when you're already working with programs
> like PyMOL and APBS.
>
> Cheers,
>
> -Michael
>
>
>
>> Good luck!
>>
>> Best,
>> Shiven
>>
>>
>>
>> On Thu, Apr 15, 2010 at 12:01 AM, J. Fleming <mol...@gm...> wrote:
>>
>>> Hi all,
>>>
>>>  I've been trying to get APBS Tools and/or APBSTools2 running for a
>>> few weeks now on multiple linux systems (RedHat and Fedora).  I've
>>> Installed MALOC and APBS Tools.  I can start up Pymol through Phenix,
>>> load a .pqr generated from a PDB2PQR server, start APBS Tools, set the
>>> grid, but when I click on Run APBS I get: "ObjectMapLoadDXFile-Error:
>>> Unable to open file!"
>>>
>>>  If I use the "pymol generated PQR" option with my .pqr even though
>>> it is a PDB2PQR server generated file APBS runs and states "X atoms
>>> did not have properties assigned" and to either remove or fix the
>>> atoms.  How can I fix them?  The "problem" residues are just regular
>>> amino acids; no split confrimations etc.  I've also tried editing out
>>> the PDBs chain name before converting to the PDB to PQR but that
>>> didn't work.
>>>
>>> Another things I've tried is using APBSTools 2 but it errors as soon
>>> as I click "set grid" with: Error: 2 ... Attribute Error: elec_radius.
>>>
>>> I really need to get this to work but I have no clue what to try next.
>>>
>>> Thanks in advance for any thoughts.
>>> -Jon
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Download Intel&#174; Parallel Studio Eval
>>> Try the new software tools for yourself. Speed compiling, find bugs
>>> proactively, and fine-tune applications for parallel performance.
>>> See why Intel Parallel Studio got high marks during beta.
>>> http://p.sf.net/sfu/intel-sw-dev
>>> _______________________________________________
>>> PyMOL-users mailing list (PyM...@li...)
>>> Info Page: https://lists.sourceforge.net/lists/listinfo/pymol-users
>>> Archives: http://www.mail-archive.com/pym...@li...
>>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Download Intel&#174; Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> PyMOL-users mailing list (PyM...@li...)
>> Info Page: https://lists.sourceforge.net/lists/listinfo/pymol-users
>> Archives: http://www.mail-archive.com/pym...@li...
>>
>
>
>
> --
> Michael Lerner, Ph.D.
> IRTA Postdoctoral Fellow
> Laboratory of Computational Biology NIH/NHLBI
> 5635 Fishers Lane, Room T909, MSC 9314
> Rockville, MD 20852 (UPS/FedEx/Reality)
> Bethesda MD 20892-9314 (USPS)
>