[CP2K:235] Re: Tutorial on cp2k

[Teodoro Laino]

May I just do  a comment, about that, based on my experience?
If the final goal is to run QMMM simulations, starting directly with  
QMMM simulations can be
quite difficult, most of the time impossible, and the risk is that  
you may give up just because
the amount of things to learn at the same time (with no documentation  
at all) is really massive.
This is a general rule obviously (not necessarily directed to Luca)..  
depending on the personal
skills and motivation you may find that starting directly with QMMM  
applies perfectly to your
learning curve for you.

In general what I suggest to people willing to try cp2k is to start  
having some confidence in the
quantum package QS, maybe just trying to reproduce full quantum  
results obtained with
other packages like  GAMESS or MOLCAS or TURBOMOLE or your favourite  
QM code..
At the same time is important to gain the same  confidence in the  
classical module FIST: running
  small/medium systems already studied with other classical codes  
(AMBER/CHARMM/NAMD/..)

Once one think to have enough control of both than QMMM comes naturally.

The idea is to learn step by step.. easier and more error proof.
Moreover, if you start QMMM simulations (with whatever package, not  
necessarily cp2k) it is
necessary that you have some experience already on both quantum  
simulation and classical
  simulations.
The risk otherwise is to build new theories based on those  
meaningless numbers
and see them published in the next number of SCIENCE or NATURE ;-)  
LOL ..

teo

On 1 Sep 2007, at 03:07, Axel wrote:

>
>
> ciao luca,
>
> On Aug 31, 8:45 am, Luca <bellu... at unisi.it> wrote:
>> I understand ....like the installation :-)
>>
>> I want to do:
>> 1-Perform MM MD for protein in solution like NAMD or ORAC MD package.
>> 2-Perform QM MD for molecules.
>> 3-Perform QM/MM MD for enzyme. QM for active site.
>> and I hope....
>> 4-"Metadynamics" for reaction coordinate in enzyme active site.
>>
>> When i will do this I can write my notes
>> and make a pseudo-tutorial for cp2k users.
>> I hope in your aid.
>
> in that case, i would suggest, that you start practicing with
> one or two of the examples posted for the demo session of
> our recent qm/mm workshop in philadelphia. see:
> http://www.cmm.upenn.edu/qmmm2007.html
>
> during trying out those we stumbled across a few useful
> improvements of the cp2k code already, and since we have
> already tried to set them up for different QM/MM code, they
> would give a foundation for a future tutorial.
>
> if i find a little more time over the weekend, i could add my
> notes on making a biosystem work with cp2k for qm/mm to
> the wiki-like part of this group.
>
> there are also a few notes in the cvs on compiling and using
> cp2k, that would be a good starting point for a tutorial and/or
> user's guide.
>
> cheers,
>    axel.
>
>> On Aug 31, 1:53 pm, Teodoro Laino <teodor... at gmail.com> wrote:
>>
>>> Ciao Luca,
>>
>>> unfortunately a tutorial is not still available. Then you may ask:
>>> how can a new user approach cp2k?
>>> Mainly with a great persistency and motivation ;-)
>>
>>> Having a look at the examples (take care that most of regtests have
>>> parameters set in order to run fast and absolutely
>>> have no sense for production runs), starting your own input and
>>> asking to the google group for  things that you don't really
>>> understand or that look weird.
>>
>>> There's some sort pg plan for a tutorial but it will take some  
>>> time..
>>> Hope though these difficulties you will enjoy CP2K ;-)
>>
>>> Teo
>>
>>> On 31 Aug 2007, at 13:47, Luca wrote:
>>
>>>> Dear cp2k users,
>>>> Where can i find tutorial or guide for cp2k?
>>>> The examples in cp2k/tests/ and "CP2K input reference" command list
>>>> are very good but not enough for
>>>> new users.
>>>> Thanks
>
>
> >