[gmx-users] how to create a dummy atom in the oplsaa force field?

[Yang Ye]

Here is an example of a simple topology of 4 atoms, fixing 1st with 
respect to no.2, 3, 4 atoms using 3out. I used the parameters from 
ffamber port from Stanford. OPLSAA FF could be a better FF for making 
dummy atoms.

;#include "ffamber99.itp"

[ defaults ]
; nbfunc        comb-rule       gen-pairs       fudgeLJ fudgeQQ
1               2               yes             0.5     0.8333

#define c6c12 6.04920e-01  3.38904e-04    
[ atomtypes ]
opls_O      O      0.0        -0.198148148 D    c6c12
opls_C      C      0.0         0.0         D    c6c12
opls_A      A     21592.4925   0.0         A    c6c12
opls_X      T     300.0        1.0         A    c6c12
opls_CL    Cl-    35.45300    -1.000       A    4.40104e-01  4.18400e-01
opls_NA1   Na+    22.98977     1.000       A    3.32840e-01  1.15897e-02
opls_K     K+     39.09830     1.000       A    4.73602e-01  1.37235e-03
opls_MG    Mg2+   24.30500     2.000       A    1.41225e-01  3.74342e+00

[ pairtypes ]
opls_X   opls_X   2 c6c12

[ moleculetype ]
;name  nrexcl
Spider  0

[ atoms ]
;nr  type        resnr    residu  atom    cgnr   charge        mass
    1 opls_O     1       SPD     O1      0
    2 opls_A     1       SPD     A1      1
    3 opls_A     1       SPD     A5      2
    4 opls_A     1       SPD     A6      3

[dummies3]
; Dummy from  from  from   funct   a          b          c
  1     2     3     4      4       0.050803   1.190606   0.044177

I have a small program to calculate parameters for type 3out dummy 
parameters based on atom coordinates. If you happened to use this type, 
I could email you in private.

Yang Ye

leafyoung81-group at yahoo.com wrote:

>Yes. It was a bit complicated to define dummy atoms
>when I did it.
>
>Here is my advices.
>
>1. Add user-defined atoms in the bottom of
>ffoplsaanb.itp. E.g.
>opls_O      O      0.0        -0.198148148 D    xxx  
>xxx
>opls_C      C      0.0         0.0         D    xxx  
>xxx
>opls_O and opls_C are defined name, variable according
>to your preference. Just remember to set mass to 0.0
>and atom type as D(for dummy)
>
>2. In the topology, use your defined type where you
>want to be.
>[ atoms ] 
>    1 opls_O     1       SPD     O1      0
>...
>   11 opls_C     1       SPD     O1      12
>
>Don't put any bonded interaction to dummies. Dummy
>atoms only have non-bonded interaction with others.
>
>3. In [ dummiesX ] section, set the parameters for the
>positioning of the dummies (refer to manual)
>
>4. There are other things needs to taken care,
>non-bonded interaction is the most important one (at
>least, for me). Do you want that dummy atom to have
>interaction with others, or with specific ones? This
>will be set according to nrexcl in [ moleculetype ]
>section and [ exclusions ] section in the topology.
>Check this with the output of gmxdump and quick MD
>tests. Until you have been sure everything is in
>control, you may start the simulation.
>
>Good luck!
>
>Yang Ye
>
>--- nanyu101 at sina.com wrote:
>
>  
>
>>Dear gmx-users,
>>
>>   I am trying to find a right way to simulate my
>>zinc enzyme. There are several methods found in the
>>references and mailing lists, such as cationic dummy
>> atom approach. But how to create a dummy atom in
>>the parameter files? The munaul is very compact, I
>>can\\\'t understand it clearly. Thanks a lot.
>>
>>all your bests,
>>______________________________________
>>¹º»ÝÆÕ²ÊÉ«È«¹¦ÄÜÒ»Ìå»úËÍÉñÆæËÄÏÀ¾«ÃÀÀñÆ·(
>>
>>    
>>
>http://ad4.sina.com.cn/sina/limeng3/mail_zhuiyu/mail_zhuiyu_20051017.html
>  
>
>>)
>>
>>
>>    
>>
>===================================================================
>  
>
>>×¢²áÐÂÀË2GÃâ·ÑÓÊÏä
>>(http://mail.sina.com.cn/chooseMode.html)
>>_______________________________________________
>>gmx-users mailing list
>>gmx-users at gromacs.org
>>http://www.gromacs.org/mailman/listinfo/gmx-users
>>Please don't post (un)subscribe requests to the
>>list. Use the
>>www interface or send it to
>>gmx-users-request at gromacs.org.
>>
>>    
>>
>
>_______________________________________________
>gmx-users mailing list
>gmx-users at gromacs.org
>http://www.gromacs.org/mailman/listinfo/gmx-users
>Please don't post (un)subscribe requests to the list. Use the 
>www interface or send it to gmx-users-request at gromacs.org.
>
>  
>

-- 
/Regards,/
Yang Ye
/Computational Biology Lab
School of Biological Sciences
Nanyang Technological University
Singapore
Tel: 6316-2884
/