[gmx-users] zero charge model- simulation time

[Justin A. Lemkul]

Juliette N. wrote:
> Hello,
> 
> I set zero charges to all atoms using the commands below:
> 
> grompp -f *.mdp -c *_em.gro -p *.top -o *_md.tpr
> 
> tpbconv –s *_md.tpr -o *-ZERO_md.tpr –zeroq
> 
> mdrun -s *-ZERO_md -o -c  -e -x -v
> 
> So before doing mdrun md I add tpbconv command and use the out put 
> *-ZERO_md.tpr for the mdrun md step.
> 
> I did not touch mdp file to exclude
> 
> coulombtype         = 
> 
> rcoulomb-switch     =  0                   
> rcoulomb            =  1.0            
> 
> Although g_Energy is giving zero values for coulomb SR and 14 terms, CPU 
> time doesnt change for the models with and without charges. 
> 
> I am just curious why run time is not changing given that electrostatics 
> is the most time consuming calculation. Do I need to remove above lines 
> from mdp? I am doubtful if electrostatic has been ignored from my model.
> 

Turning off charges will not improve performance.  The MD routine will still 
calculate Coulombic interactions, but it will just waste time producing all 
zeros.  The only difference you might be able to see would be PME vs. cutoff 
(the default coulombtype setting), with the latter being much faster (but also 
much less accurate).

The only way to exclude these interactions is to use energygrp_excl, but that 
will exclude all nonbonded interactions, not just the electrostatics terms.

-Justin

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================