[gmx-users] Document about file input format to GROMACS program?

[root]

Dear GROMACS user:
We downloaded the AMBER-03* potential parameters files
(ffamber_v3.3.1-doc.tar.gz) linked with the page
http://folding.stanford.edu/ffamber/index.html#validation. 
We tried to simulate the Dickerson dodecamer using the 3.3.1 gromacs
version.
Our simulation crash when we run the grompp program. A (nonexistent)
error in the atom numeration of the topology is indicated by the
program. 
Previously, we run the pdb2gmx and the genbox programs solving notation
errors. Finally .top and coordinate files are generated without
(apparently) errors.
Analyzing the procedure, we observed that the format of the input files
of the AMBER-03* potential (.rtp,.itp, etc.) frequently do not coincide
with the input files included in the share/top folder of gromacs.
For example the ffamber03.rtp file describe the bonds:
....
[ bonds ]
   H5T   O5'
   O5'   C5'
   C5'  H5'1
   C5'  H5'2
   C5'   C4'
   C4'   H4'
....
omitting the indication of the bond type, how for example occurs in
ffG43a1.rtp
.....
 [ bonds ]
    N     H    gb_2    
    N    CA    gb_20   
   CA     C    gb_26   
    C     O    gb_4    
.......
The list of differences is long, and this is only one example.

The questions are:

a) There are any document indicating the format of the file inputs for
the gromacs programs?
The user manual 3.3 include little or not at all about this.

b) Is the cause of the error these format differences?

Thanks in advance

Mario Campo and Griselda Corral
Fac. de Cs. Exactas y Naturales
UNLPam