[gmx-users] Re: Regarding Gromacs problem
Justin A. Lemkul
jalemkul at vt.edu
Tue Aug 25 15:33:18 CEST 2009
Please do not send private emails asking for help. I've said this dozens of
times; it is much more beneficial to the community as a whole to post the
gmx-users list, where discussions are archived for future use. You also have a
better chance of reaching someone who knows how to solve your problem. I know I
respond a lot, but I don't always have an answer, nor do I advertise myself as a
private tutor. Contribution to the list is voluntary, when I've got a few moments.
Your system is blowing up, a fairly common problem. See here for general tips:
rituraj purohit wrote:
> Hello Justin..
> I am getting a problem during dynamics(position restrain) of pr.mdp,
> at first i given dt = 0.002 and for 200ps (because protein is very big
> having 1480 aa = 11058 atoms) shows an error called table size at 1.6
> ps. Then i changed it dt = 0.001 for 200ps, but again i had a same
> error at ~45 ps.
> So finally i giving dt = 0.0001 for 200 ps, now its moving fine..but
> steps become 20 lacks :(
> 1) The correction which i done , is it ok?
You're probably just prolonging the inevitable; the system is not stable.
> 2) Is error occurred (with dt = 0.002 & 0.001) due to big size of protein?
No; it is due to some physically unrealistic aspect of the model physics
(unresolvable clashes, inappropriate parameters). Size should not be a problem;
people are simulating viral capsids nowadays.
> 3) To do dynamics for such a big protein is rel event?
That's up to you.
> 4) Is 200 ps enough for this protein? is their any relation between
> protein size and pre-equilibrium period ?
There is no standard rule, really. To justify equilibration, your observables
have to have converged (i.e., temperature, pressure, density, RMSD, etc).
> Please help me in this..
Justin A. Lemkul
ICTAS Doctoral Scholar
Department of Biochemistry
jalemkul[at]vt.edu | (540) 231-9080
More information about the gromacs.org_gmx-users