[gmx-users] Fatal Equilibration Errors

[Justin A. Lemkul]

Nancy wrote:
> Hello,
> 
> I have successfully minimised and equilibrated ethylene glycol in a 
> water box.  I have noticed that there seem to be no hydrogen bonds 
> between the solute and solvent, but there are hydrogen bonds forming and 
> breaking between solvent molecules.  Is this a normal behavior during 
> minimsation and equilibration?
> 

That will depend on whether or not you are position-restraining your ethylene 
glycol, and whether or not your parameters are consistent with the derivation of 
the force field.

<snip>

> Steepest Descents converged to machine precision in 7007 steps,
> but did not reach the requested Fmax < 10.
> Potential Energy  = -3.3459230e+04
> Maximum force     =  4.9560604e+01 on atom 4
> Norm of force     =  3.4399371e+00
> ===============================================================
> 
> As I believe these forces to be acceptable, I proceed to equilibration:
> 

These values appear to be reasonable, yes.

> $ grompp -f nvt.mdp -c em.gro -p glycerol.top -o nvt.tpr
> 
> where my "nvt.mdp" file is:

> rvdw        = 1.3

As I've said before, this value of rvdw is not what is expected when using the 
G53a6 parameter set.  Unless you've got a good reason for altering the original 
force field specifications, generally this is not a good idea.

> tcoupl        = V-rescale
> tc-grps        = GOL SOL

How many glycerol molecules do you have?  If it is only a small fraction of the 
system, this treatment of thermostats is likely not appropriate.  See here:

http://oldwiki.gromacs.org/index.php/thermostats

<snip>

> ===============================================================
> Step 0, time 0 (ps)  LINCS WARNING

The procedure for solving a LINCS warning is always the same.  Identify where 
things start to fall apart, then verify that your topology is correct, and your 
.mdp parameters are reasonable.  I've made a few comments on .mdp parameters 
above.  As for the topology, that still remains a mystery.  Having never used 
topolbuild, I don't know how it assigns parameters, so it may work just fine. 
But in general, this is always part of the diagnosis.

-Justin

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

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

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