[gmx-users] Re: AMBER force fields for ATP

[Thomas Piggot]

No intrusion, but I have forwarded this to the GROMACS mailing list as 
all correspondence is best kept there. It gives you have a much greater 
chance of getting (better) help.

The information needed for the parameters is best obtained from the 
supporting information given with the Carlson paper. The first thing to 
do is to add an appropriate entry to the .rtp file. For the O3 atom you 
need to give a new atom type (for example I have it as amber99_68). You 
then need to add this new amber atom type to the nb.itp and .atp files. 
The non-bonded parameters of the new O3 atom type are the same as O2, so 
to add this new atom type to the nb.itp file is trivial, and the same is 
true for the .atp file.

For the bonded parameters you need to edit the bon.itp file. Firstly you 
need to add an entry into the [bondtypes] section for the O3 P bond. 
Then you need to add entries into the [angletypes] for the O3 P O3 and 
O3 P OS angles. Like the non-bonded parameters these are trivial as they 
are by analogy to the O2 atom type interactions. For the [angletypes] 
you also need to change the default P OS P values to those given in the 
supplementary info of the Carlson paper. This could cause a problem if 
you have another molecule in your system which uses this angle and you 
don't want it to use the new angle parameters. I did not and so did not 
worry about changing this value. If you do use this angle in another 
molecule (and the default values of some of the P dihedrals which will 
be changed next) then it would be easier to make an .itp file (where you 
can just override the defualt values by having the new values in the 
.itp), or introduce a new atom type for the P atom and use this in all 
the bonds/angles/dihedrals/non-bonded interactions. Finally you need to 
add in the new proper dihedrals from the supplementary info into the 
[dihedraltpyes] section. To turn these parameters given into the correct 
RB format you need to use the equations in the GROMACS manual (4.63 of 
the GROMACS 3 manual/4.64 GROMACS 4 manual). As with the [angletypes] 
you need to alter the default CT OS P OS dihedral to the new value from 
the paper.

Hope this all makes sense and is all correct (I did this quite a while 
ago!).

Cheers

Tom

Alice Chang wrote:
> Dear Tom,
> 
> I hope this is not an intrusion, but I saw your answer on the gmx-users 
> mailing list about finding ATP parameters for use in GROMACS. I am 
> trying to solve the same problem, but without his luck! I've examined 
> the PREP and FRCMOD files, as well as the relevant Carlson paper and 
> ffamber files. I've been comparing the amber .itp, .atp, and .rtp files, 
> but I'm not sure what to use for the radius and epsilon values, or how 
> to parametrize the P-OS-P bond. I was wondering, if you don't mind, 
> could you point me in the right direction?
> 
> Thanks very much,
> Alice Chang
> Liao Research Group
> Columbia University

-- 
Thomas Piggot
University of Bristol, UK.