[gmx-users] Question about the dihedrals in top file using OPLS force field in GROMACS

[Mark Abraham]

Yanmei Song wrote:
> [ dihedrals ]
> ; ai  aj  ak  al  fu    c0, c1, m, ...
>   1   2   5   6   1    0.0    3.77   3    0.0    3.77   3 ; dih   CAA
>  CBF  OAW  CBH
>   2   5   6   9   1    0.0    3.77   3    0.0    3.77   3 ; dih   CBF
>  OAW  CBH  OAY
>   5   6   9  10   1    0.0    3.77   3    0.0    3.77   3 ; dih   OAW
>  CBH  OAY  CBJ
>   6   9  10  13   1    0.0    3.77   3    0.0    3.77   3 ; dih   CBH
>  OAY  CBJ  OBA
>   9  10  13  14   1    0.0    3.77   3    0.0    3.77   3 ; dih   OAY
>  CBJ  OBA  CBL
> 
> Sorry to bother you so much times. Above is the real parameter I used
> in my top file which probably specified the dihedrals as periodic
> (function 1) . But I also used OPLS force field which should read
> different parameters as RB. So when the MD runs, does it just read
> this parameters in the periodic dihedrals as a RB parameter?

The fifth column determines the functional type used, per the table in 
chapter 5 to which I referred you. One can mix and match these at will - 
apparently OPLS routinely uses R-B. That doesn't stop you using periodic 
ones as well.

> If my top
> is in a wrong form, how can md runs well and give a good result?

Maybe it does for the observable and time scale you've looked at. One 
can't mix force fields and a priori expect the results to be correct. 
Extensive testing may reveal an acceptable result. See 
http://wiki.gromacs.org/index.php/Parameterization

Mark