[gmx-users] Re: Distance restraints not working
Justin A. Lemkul
jalemkul at vt.edu
Wed May 23 20:40:28 CEST 2012
On 5/23/12 8:26 PM, tdgrant1 wrote:
> There are a few possibilities.
>> 1. Numbering in [distance_restraints] is not based on .gro numbering. The
>> numbers supplied are only relevant within the [moleculetype] to which they
>> belong, much like [position_restraints]. Also make sure that the
>> [distance_restraints] block is in the correct location in the topology.
>> 2. You've omitted the dihre_fc keyword, though it is set to 1000 by
>> Check the mdout.mdp file or your .tpr file to ensure that you have a
>> value here. Setting values of fac that high are unnecessary.
> Thanks for your help. I have it working now. It actually turned out to be a
> I have an unrelated question though. After running the temperature and
> pressure equilibration, grompp tells me my md run will produce ~150 GB of
> data. I'd really like to avoid such an enormous amount of data. The system
> is ~300,000 atoms and I will be running 20 ns of simulation (~3 days on our
> cluster according to the equilibration output). Can I simply change the
> output control from saving the various files at every 2 ps to every, say, 20
> ps? I realize it's a large jump compared to 2 ps, but 150 GB is a lot of
> data. Will changing the output control so drastically adversely affect the
> simulation in any way?
Saving output has no bearing on the simulation result. Writing every 2 ps is
probably drastic overkill for most situations. You should also evaluate what it
is you need to save, and in what precision. For instance, if you only need
coordinate information, most analysis can be conducted on .xtc files, which take
less space. If you don't need velocities or forces, you can set nstvout and
nstfout to zero. If you only care about .xtc precision for coordinates, you can
eliminate the .trr file altogether by also setting nstxout to zero.
Justin A. Lemkul, Ph.D.
Department of Biochemistry
jalemkul[at]vt.edu | (540) 231-9080
More information about the gromacs.org_gmx-users