[gmx-users] How to set up dihedrals

[Mark Abraham]

Thomas Schlesier wrote:
> Dear all,
> i'm making a .rtp for a sugar for the OPLS-AA force-field. To start i
> made a topology with the PRODRG server, so i have a list with all bonds,
> angles and dihedrals, but the problem is, that i'm missing the unpolar
> hydrogens. How to set up the bonds and angles entrys is no problem, but
> with the dihedrals i have problems.
> 
> If i have for example the the following molecule (should look like a
> double Y):
> 
> L1\   /R1
>    L-R
> L2/   \R2
> 
> (L1 and L2 bounded to L, then L to R, and R is bounded to R1 and R2)
> In this case i have the dihedral between L and R, but i have 4
> possibilities to set it up:
> L1-L-R-R1
> L1-L-R-R2
> L2-L-R-R1
> L2-L-R-R2
> So my problem is how to choose which possibility should i use. If L1, L2
> and R1, R2 are equal there is no problem, because then the 4
> possibilities are all equal.

The only way to be sure is to read the primary literature for OPLS/AA. 
You might also look at some other .rtp file entries that have the 
topology you're considering.

Mark