[CP2K:2029] Re: visual approach to construction of QMMM input

[Laino Teodoro]

Hi Peter,

I understand your point and somehow I share your view.. not fully  
though..
The major issue is that a candidate GUI should not map the input  
structure 1<->1, but should
be smarter than just providing the possibility to edit all possible  
fields of the input.

Anyway.. this was just a "feature request" as a user ;-)

In addition, since pymol offers as well a molecular editor this would  
make the whole thing (pymol+input applet) extremely
powerful and viral in terms of possible diffusion of CP2K :-) (on the  
other side a review of all default values will be another
essential key-point).

Ok.. maybe I'm still dreaming..  :-)
Thanks anyway for the scripts and please keep-us updated in case of  
further script/applet development related to
CP2K on pymol.

Cheers
Teo

On 21 Apr 2009, at 17:41, Peter wrote:

>
> Hello Teodoro!
>
> Thank you for your interest : )
>
>> do you think would it be easily feasible
>> to have a full applet (with input field structure parsed from an
>> up-2-date xml cp2k input tree) in order to allow the creation of the
>> whole CP2K input graphically?
>
> Well, it is definitely possible and probably not very hard to
> implement. But do you think it will be easier to find some parameter
> input field in GUI than write it in plain text? Especially considering
> the number of  parameters...
> I had some experience with GAMESS graphical input generators and IMHO
> they are useless. But I'm definitely a "command shell guy" : )
> Anyway, I would participate in integration of CP2K and PyMol with
> great pleasure.
>
> Peter
>
> On Apr 21, 6:55 pm, Teodoro Laino <teodor... at gmail.com> wrote:
>> Hi Peter,
>>
>> thanks for posting this useful info!
>> I'm pretty sure this will be definitely useful.
>>
>> I have a feature request! :-p
>>
>> Im not an expert of pymol, but do you think would it be easily  
>> feasible
>> to have a full applet (with input field structure parsed from an
>> up-2-date xml cp2k input tree) in order to allow the creation of the
>> whole CP2K input graphically?
>> This, combined with the graphical handling of the molecular  
>> system, may
>> be really an interesting feature.
>>
>> I'm ready to bet that this would have even more appeal ;-)
>> Teo
>>
>> Peter wrote:
>>> Dear CP2K users,
>>
>>> Recently I've written a simple script, which could be helpful to  
>>> those
>>> of us, who deals with large QMMM systems.
>>> The script is intended to be used from PyMol (http://www.pymol.org/)
>>> molecular visualization system and allows one to generate &QM_KIND,
>>> &LINK and &CELL parts of CP2K input easily. Given a molecular object
>>> name and QM part selection name (which can be defined using either
>>> approach available in PyMol) it generates output file with
>>> corresponding input groups.
>>> To use the script:
>>>   1. Download it (http://groups.google.com/group/cp2k/web/gen-cp2k-
>>> qmmm.py)
>>>   2. Start PyMol and issue the following command from it's shell:
>>>      >  run /path/to/script.py
>>>   3. Use it:
>>>     > gen_cp2k_qmmm filename, qm_part_selection_name,
>>> the_whole_system_object
>>> The output will be saved to file "filename".
>>> N.B.: The file will be overwritten!
>>
>>> Another PyMol script which you may find handy is a multiframe XYZ  
>>> file
>>> (trajectory) loader (http://groups.google.com/group/cp2k/web/
>>> load_xyz_traj.py).
>>> "Installation procedure" is pretty the same.
>>> Usage:
>>>  > load_xyz_traj file.xyz, target_object [, stride]
>>
>>> N.B.: By default PyMol reorders atoms on molecule load. This PyMol
>>> behavior may prevent correct loading of XYZ trajectories. To
>>> workaround this issue add the following lines to ~/.pymolrc file
>>> (create it if needed):
>>> set retain_order, 1
>>> set pdb_retain_ids, 1
>>
>>> Hopefully these simple scripts will help people familiar with  
>>> PyMol in
>>> doing biostuff with CP2K.
>>
>>> Best regards,
>>> Peter
> >