[CP2K:2545] Constraint warning with AMBER PDB file

[marco.stenta]

Dear Teodoro,
I'm running QM/MM cp2k on a pretty big system.
I'm using amber topology so I need a way to SHAKE the water molecules
I understand I do it in the &MOTION section with something like that, using 
molecule as, without a psf I cannot specify the molname 

  &CONSTRAINT
    &G3X3
      MOLECULE                     31
      DISTANCES                    1.7958  1.7958  2.85444
      ATOMS                        1 2 3
      EXCLUDE_QM
    &END G3X3
  &END CONSTRAINT

Here 31 is the molecule kind number  of the fist water molecule as guessed 
from the output of a tentative run with verbose output: I get the molecule 
kind name is MOL30 and the molecule kind number is 31: is it right?

    31. Molecule kind: MOL30                     Number of 
atoms:              3
                     Atom         Atomic kind name
                        1                       OW
                        2                       HW
                        3                       HW

        The name was automatically generated: T
        Number of molecules:      1
        Molecule list:               31
        Number of bonds:            3



The problem is that I have 47883 water molecules: do I have to repeat this 
CONSTRAINT  section (or the &G3X3 subsection ) 50 thousand time? the input 
would be pretty big.
Is there a clever way to SHAKE water molecules in my case?

you write in your previous answer 
MOLECULE 1..15

but from the manual I understand MOLECULE takes only one integer: what does 
your line mean? can I specify for a range instead then for a single 
molecule?

Is it possible to have the code recognizing a MOLNAME also when one uses an 
amber topology?
I am forced to use amber topology as I did not find the ambertocharmm 
converter (leap2fist ?) anymore and I have to shake water, of course.

More in general: what do you mean by "molecule"?
each residue of a protein is assigned to a "molecule": what is the  meaning 
of that?
TOPOLOGY > PARA_RES<http://cp2k.berlios.de/manual/CP2K_INPUT/FORCE_EVAL/SUBSYS/TOPOLOGY.html#list_PARA_RES>is set to TRUE
what if I set it to FALSE?
I get the term molecule has nothing to do with its chemical meaning, but my 
protein part is composed by more than one sub-units: does that conflicts 
with PARA_RES<http://cp2k.berlios.de/manual/CP2K_INPUT/FORCE_EVAL/SUBSYS/TOPOLOGY.html#list_PARA_RES>FALSE?
Then I get some strange molecules as well like 
     1. Molecule kind: _QM_MOL0                  Number of atoms:          
10657
this is not the QM part, nor the QM + link and I am lost 

thanks for the help
m


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20110621/8ebeda94/attachment.htm>