Basis Sets and Kinds and Amber

[Rolf David]

Hi all,

I'm working on a QM/MM calculation
I use as coord/topology/FFparamaters a amber12 restart/prmtop.

Everything is nice, except a :

*** Unknown element for KIND <OP>. This problem can be fixed specifying *** 
*** properly elements in PDB or specifying a KIND section or getting in *** 
*** touch with one of the developers! ***

Of course I have OP Kind in my MM and QM parts (in amber: name OP, type OP, 
element 8)
(Also another SM Kind only in my MM part, i did the same as OP)

So, I put a :

&KIND OP
 ELEMENT O
&END KIND
 
In the Subsys section, also I have :

&KIND O
 BASIS_SET 6-31Gxx
 POTENTIAL ALL
&END KIND

So everything went smoothly according to plan, and suddenly, I decided to 
change a part:

&KIND OP
 ELEMENT O
 BASIS_SET 6-31++Gxx
 POTENTIAL ALL
&END KIND

and by reading the output all oxygens (O Element) are treated as 6-31Gxx
I change the order (&Kind O after &Kind OP) no luck.

How kind/type/element from amber interact with CP2K ?

I suppose the &KIND in the &SUBSYS are linked to the &QM_KIND in the &QMMM 
part, only for QMMM atoms ?

KIND OP exists only amber parm/rst file, so I needed to declare the element 
-> For the MM part.

But since in my &QMMM part, I only have &QM_KIND O for all my oxygens, 
putting a basis/potential for OP has no effect so I have to declare a 
&QM_KIND OP.
By putting &QM_KIND OP, the &KIND OP basis part worked.

What about atom type (amber type, amber name, so complicated...) like for 
TIP3P water, O is the name, and OW is the type
in the &QMMM part I have :

&MM_KIND OW
 RADIUS 0.78
&END MM_KIND

It will be applied only to MM atoms, of amber type OW ?

&KIND correspond to amber name or amber type or both or whatever it feels 
like it ?

Confirmations or corrections will be &KINDly appreciated.

Thanks.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20150224/ce69de6c/attachment.htm>