Re: [AMBER] -nan values.

From: Jason Swails <>
Date: Mon, 25 Aug 2014 13:27:02 -0400

On Tue, 2014-08-26 at 00:36 +0800, Dhananjay wrote:
> I am relatively new to MMPBSA calculations and trying to run mmpbsa for
> protein-protein interactions with 'multiple trajectory protocol'.
> I have MD trajectories for complex, ligand and receptor.
> Before MD run, I make sure that the # of atoms(complex.pdb) = # of
> atoms(receptor.pdb+ligand.pdb).
> I used following command to run mmpbsa:
> mpirun -np 8 -O -i -cp
> ../preparation/complex-vac1.prmtop -sp complex-bondi.prmtop -rp
> -lp -srp receptor-bondi.prmtop -slp
> ligand-bondi.prmtop -y
> ../mult_traj_prot/complex/04_MD_run/protein_prod_nvt.mdcrd -yr
> ../mult_traj_prot/receptor/04_MD_run/protein_prod_nvt.mdcrd -yl
> ../mult_traj_prot/ligand/04_MD_run/protein_prod_nvt.mdcrd
> Although, I am getting "MM/PBSA processing is done !" ther are still "-nan"
> values in the output.
> How can I fix this problem ?
> Is my protein-protein complex system is too large ?
> OR I missed something here ?

These problems almost always indicate an incompatibility between the
generated trajectory files and the prmtop files you fed to

Visualize the trajectories that were generated
(_MMPBSA_receptor.mdcrd.0, _MMPBSA_ligand.mdcrd.0, ...etc) with the
corresponding prmtop files (,, etc.) My bet
is that the structures look completely messed up.

Oftentimes, this is caused because the default strip_mask either strips
out too many items (if you have a metal ion or water bound somewhere,
for instance), or it is not stripping enough stuff.

The default value of strip_mask is shown in the Amber 14 manual in the chapter.


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Mon Aug 25 2014 - 10:30:02 PDT
Custom Search