[gmx-users] Reduction in dihedral types after grompp

[francesco oteri]

Dear gromacs users,
I would like to know whether in grompp there is a process
that in someway reduces the number of dihedral stored in tpr.?

I have the attacched files.
When I use the first .top file I obtain a smaller .tpr.
The only difference between the two files are the charge and epsilon
paramaters of [ atoms ] section and the constant of some dihedral
angles. However, the names of the atoms are exactly the same,
only the constant changed.

I investigated through gmxdump what is the difference between the
two generated .tpr and I found that the number of PDIHS and IDIHS
entry is different for reft.top_0.tpr and reft.top_1.tpr:

196 and 274 PDIHS reft.top_0.tpr vs reft.top_1.tpr
10           18 IDIHS  reft.top_0.tpr vs reft.top_1.tpr

to generate the files, I've done:
grompp -p reft.top_0.top -n index.ndx -f reft.mdp -c 04md.gro
-o reft.top_0.tpr
grompp -p reft.top_1.top -n index.ndx -f reft.mdp -c 04md.gro -o
reft.top_1.tpr

Can you tell me why this phenomenon happens?

Francesco

 reft.top_1.tpr<https://docs.google.com/file/d/0B1ktZ-u2OHFAZXQtalc2WEFjTUk/edit>

 reft.top_1.top<https://docs.google.com/file/d/0B1ktZ-u2OHFAZTVPNGVIdTg3eG8/edit>

 reft.top_0.tpr<https://docs.google.com/file/d/0B1ktZ-u2OHFATjhBNWJVY3JGNHM/edit>

 reft.top_0.top<https://docs.google.com/file/d/0B1ktZ-u2OHFAUlVQOWo4VlVCS3c/edit>

 reft.mdp <https://docs.google.com/file/d/0B1ktZ-u2OHFAMzR0VHlwSXc3elE/edit>

 index.ndx<https://docs.google.com/file/d/0B1ktZ-u2OHFALVdDR19sYWFyVDg/edit>

 04md.gro <https://docs.google.com/file/d/0B1ktZ-u2OHFAWm54aEgzMHU3Y0U/edit>