[gmx-users] coulombtype selection

[Jian Zou]

Hi all,

I'm doing simulation of dna in water.

At first I choose for mdp option
nstlist = 5
rlist = 2.0
comlombtype = PME
rcoulomb = 2.0
rvdw = 2.0
fourierspacing = 0.1
optimize_fft = yes
But the simulation takes a long time (5 more days for 1ns run)

Does PME is necessary for my charged system?
how about change to comlombtype = cutoff and reduce the cutoff?
How much will the change PME->cutoff reduce the accuracy?

Thanks a lot in advance.


Regards,

Jian Zou