[gmx-users] Error when performing MD with heavy atoms frozen

[Justin Lemkul]

On 6/15/17 11:57 AM, Kai Hu wrote:
> Hi there,
> I am new to GROMACS, went through Lysozeme in Water tutorial and everything
> runs smoothly.
> I then tried to immobilize all heavy atoms (except H) in md.mdp file by
> specifying freezegrps               = Protein-H       ; Freeze all heavy
> atoms
> freezedim                = Y Y Y             ; Freeze all dimensions
> towards the end of the file. I previous generated a index.ndx file using
> gmx make_ndx command.
> 
> This time, a fatal error would show up, saying:
> Fatal error:
> Bond length not finite.
> 
> Can anyone tell me briefly what might be wrong that leads to such error?
> Note, if I remove these two commands in md.mdp file, no error would pop up,
> it seems something must wrong with the freeze commands.
> 

Freezing is an extremely artificial perturbation and if you're applying a 
barostat, you'll get spurious contributions to the virial, as noted in the 
manual.  Application of freezing usually requires corresponding energygrp_excl.

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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