VMD-L Mailing List

From: Diego Gomes (diego.enry_at_gmail.com)
Date: Thu Apr 11 2024 - 11:38:36 CDT

Never had this problem before.

The message you've sent indicates VMD is reading Surf output ( .tri file in
/tmp).
For some weird reason it is just not displaying the surface.
Info) This surface is made with SURF from UNC-Chapel Hill. The reference
is:

Info) A. Varshney, F. P. Brooks, W. V. Wright, Linearly Scalable Computation

Info) of Smooth Molecular Surfaces, IEEE Comp. Graphics and Applications,

Info) v. 14 (1994) pp. 19-25.

Info) Reading Surf geometry output file...

Info) Read Surf output file, processing geometry...

Info) Done.

Can you:
* check if QuickSurf works?
* manually write to the /tmp folder? If for not, you may change the
environment variable VMDTMPDIR

No matter the result, consider trying NanoShapper and/or MSMS, they're
faster than Surf.

On Thu, Apr 11, 2024 at 10:31 AM Dmitry Suplatov <genesup_at_gmail.com> wrote:

> Hi Diego,
>
> Thanks for the feedback.
>
> This problem persists for any proteins of any size. So the particular
> protein is not the issue. Perhaps, the problem is with my access to the tmp
> folder, I don’t know.
>
> So, did you experience anything similar when installing vmd to a
> non-default location?
>
> Thanks again,
> Dmitry
>
>
> On Thu 11. Apr 2024 at 17:14, Diego Gomes <diego.enry_at_gmail.com> wrote:
>
>> Hi Dmitry,
>> this is unusual for a molecule with only ~6800 atoms. I can help you
>> debug if you share the molecule.
>>
>> The temporary file location is hardcoded in C++, you may change at will
>> but you will need to recompile VMD.
>>
>> Friendly,
>> Diego.
>>
>>
>> On Thu, Apr 11, 2024 at 8:42 AM Dmitry Suplatov <genesup_at_gmail.com>
>> wrote:
>>
>>> Hello,
>>>
>>>
>>>
>>> I am having problems with ‘Surf’ as a “Drawing method” for proteins in
>>> VMD.
>>>
>>>
>>>
>>> So, I installed VMD 1.9.3 from this package -
>>> vmd-1.9.3.bin.LINUXAMD64-CUDA8-OptiX4-OSPRay111p1.opengl.tar.gz. However,
>>> with VMD 1.9.4a I have the same problem.
>>>
>>>
>>> I used a default sequence of commands to deploy the package. However, it
>>> is important to note that I *installed it to my local folder*, not the
>>> global one.
>>>
>>>
>>>
>>> tar xzf vmd-1.9.3.bin.LINUXAMD64-CUDA8-OptiX4-OSPRay111p1.opengl.tar.gz
>>> cd vmd-1.9.3/
>>> mcedit configure
>>> -> # Set these variables at the top of the script:
>>> -> #$install_bin_dir="$ENV{'HOME'}/vmd/bin";
>>> -> #$install_library_dir="$ENV{'HOME'}/vmd/lib/$install_name";
>>> ./configure
>>> cd src/
>>> make install
>>> export PATH=$HOME/vmd/bin/:$PATH
>>>
>>> Now I can run VMD using the ‘vmd’ command. When I load a pdb file and
>>> select ‘Surf’ as the ‘Drawing method’ the following happens - the
>>> 3D-viewer remains absolutely empty, I see this output in the bash
>>> command line of Linux, from which the vmd was launched:
>>>
>>>
>>>
>>> vmd > Reading 6882 atoms..done
>>>
>>> Probe radius = 1.400
>>>
>>> Constructing solvent-accessible surface ..
>>>
>>> Info) This surface is made with SURF from UNC-Chapel Hill. The
>>> reference is:
>>>
>>> Info) A. Varshney, F. P. Brooks, W. V. Wright, Linearly Scalable
>>> Computation
>>>
>>> Info) of Smooth Molecular Surfaces, IEEE Comp. Graphics and Applications,
>>>
>>> Info) v. 14 (1994) pp. 19-25.
>>>
>>> Info) Reading Surf geometry output file...
>>>
>>> Info) Read Surf output file, processing geometry...
>>>
>>> Info) Done.
>>>
>>>
>>>
>>> The next immediate action that I did was to investigate if the ‘surf’
>>> binary is correctly installed and set. So, while VMD was still running, I
>>> located the ‘surf_LINUXAMD64’ binary and renamed it:
>>>
>>>
>>>
>>> mv lib/vmd/surf_LINUXAMD64 lib/vmd/surf_LINUXAMD64_bak
>>>
>>>
>>>
>>> Then, I again selected ‘Surf’ as the ‘Drawing method’, and saw the
>>> following in the command line:
>>>
>>>
>>>
>>> vmd > sh: line 1: [my home folder]/vmd/lib/vmd/surf_LINUXAMD64: No such
>>> file or directory
>>>
>>> ERROR) Cannot read SURF output file: /tmp/vmdsurf.u1713953541.664.in.tri
>>>
>>> Info) Done.
>>>
>>>
>>>
>>> So, this tells me that the ‘surf’ binary exists and VMD is recognizing
>>> it (i.e. it throws an error when such binary is renamed, and no error when
>>> its in place). But for unknown reason the ‘Surf’ representation still does
>>> not work.
>>>
>>>
>>>
>>> *Questions:*
>>>
>>> 1. *Did anyone experience this and knows how to solve this?*
>>> 2. *Perhaps, the problem is with the temporary folder that is used
>>> to exchange data between VMD and Surf binary. Is there any way to set this
>>> folder from /tmp to a specific location?*
>>>
>>>
>>>
>>> Thanks and all have a great day
>>>
>>> Dmitry
>>>
>>>
>>
>> --
>> Diego Enry B. Gomes, PhD
>> Department of Physics at Auburn University &
>> NIH Center for Macromolecular Modeling and Visualization
>> Leach Science Center - Ste. 3182 - Auburn, AL
>> dgomes_at_auburn.edu
>>
> 

-- 
Diego Enry B. Gomes, PhD
Department of Physics at Auburn University &
NIH Center for Macromolecular Modeling and Visualization
Leach Science Center - Ste. 3182 - Auburn, AL
dgomes_at_auburn.edu