[gmx-users] oplsaa force field

[Mark Abraham]

The answers to both of these questions are in the comments in
src/kernel/toppush.c, in function push_at(). I found them grepping
src/*/*.[ch] for scanf (the C function you'd use to read up these files)
and looking for topology-related filenames...

> I'd like to use oplsaa FF format to build my own force field for my
> project. After I finished it, I can use pdb2gmx. but not grompp. The
> problem is from my **nb.itp file. I have specified some unusual (not
> follow the combination rule) nonbonded interaction parameters after
> [ atomtypes ] section as follows:
>
>  [ nonbond_params ]
> ; i    j func          sigma      epsilon
> CT    CT  1	       3.80000e-01  2.09200e-01
>
> When I run grompp, the program complains that atom type CT is not found.
> If I change CT to opls_182, then it works fine. However, when I check
> oplsaabon.itp, CT is used, but not opls_182. Could anyone tell me why I
> can't use atom type in *nb.itp? How can I fix it? If opls_182 is defined
> as atom type, then how does gromacs recognize CT is **bon.itp?

OPLS/AA appears to use different names for bonded and non-bonded atom types.
 See line 172 of the above file.

> By the way, my question relates to my previous question. Unfortunately,
> no one answered it. I really appreciate someone can help me. The
> following is my previous email:
>
> In ffoplsaanb.itp, there are 8 columns.
> ; name  bond_type    mass    charge   ptype          sigma      epsilon
>  opls_001   C   6      12.01100     0.500       A    3.75000e-01
> 4.39320e-01 ; SIG
>
> In ffG53a5nb.itp, there are 7 columns.
> [ atomtypes ]
> ;name  at.num   mass      charge ptype       c6           c12
>     O    8      0.000      0.000     A  0.0022619536       1e-06
>
> Could anyone tell me how gromacs knows to read 8 columns for oplsaa FF,
> but 7 columns for gromos FF? Thanks!

The algorithm is described just below in that file too.

Mark