[gmx-users] Re2: problem wiht topology and gmxdump...

[sgaray at fbcb.unl.edu.ar]

Dear Dr. Feenstra, I have some other questions about your answers: 

>IIRC, you can only have one global [nonbond_params] directive.
>So, if you have one in each of your .itp files, I think it won't work.

I have only one file with the [nonbond_params]. 

>These are in the .tpr file, as you saw using gmxdump. I'm not sure 
>exactly how they are organized, but in general there should be 
>'general' parameters for different atom types, and then some 
>specific parameters for some pairs of atom types. gmxdump may 
>list all pairs of atom types explicitly.

I´ve been seen the gmxdump outfile in detail. I have 11 atomtypes 
defined. I found 121 LJ functypes (how can calculate this quantity: 
121. Is that all posible variations,n^2? ). The file just only shows all 
pairs of LJ14 interactions but not the pairs of LJ interactions. Is it 
wrong?. I understand that in the topology file shouldn´t appear all LJ 
interactions (atom by atom) because that depend the position of the 
particles, but I would like to know wich LJ interaction is used by 
gromacs between each pair of atomtypes. The LJ functypes which 
are described by the gmxdump_outfile don´t say anything about 
which atoms are involved. Example:  

functype [115]=LJ  C6   2.52514891e-03	C12	2.11139763e-06  

I don´t know which atomtypes are involved in that interaction.
Could you tell me how gromacs build my 121 LJ functypes ?

Thank you in advance.

Sergio 

Sergio Garay
Bioquimico
J.T.P. Practica Profesional
Facultad de Bioquimica y Cs. Biologicas
Universidad Nacional del Litoral
Santa Fe (3000)- Paraje "El Pozo"
Argentina