GMX bar fatal error

User discussions
1

GROMACS version: 2020.4
GROMACS modification: No

Hello!

I have trouble running my free energy analysis of my DNA structure at different solvent conditions. During the gmx pdb2gmx run, I include -merge all option to my DNA structure so that I can analyze the whole DNA structure, not the individual chains alone. During the NPT ensemble, I encountered this error:
################################################################
Program: gmx mdrun, version 2020.4
Source file: src/gromacs/pbcutil/mshift.cpp (line 904)

Fatal error:
There are inconsistent shifts over periodic boundaries in a molecule type
consisting of 758 atoms. The longest distance involved in such interactions is
7.374 nm which is above half the box length. This molecule type consists of
muliple parts, e.g. monomers, that are connected by interactions that are not
chemical bonds, e.g. restraints. Such systems can not be treated. The only
solution is increasing the box size.

For more information and tips for troubleshooting, please check the GROMACS
website at http://www.gromacs.org/Documentation/Errors
########################################################################

I already included -rdd option and table-extension options on my simulations. What is the main cause of this error? I received this error when my PBC was 1.0 nm and also I received this kind of error using 5 nm of PBC.

2

Hi,

Could you provide more info on your system. Is your DNA double stranded? and is the DNA helix 7.4 nm long? How did you build the box around?

Just a general suggestion to get better use of the forum, use a post title that most describe your problem. Here the problem is in gmx mdrun and in gmx bar.

Best regards
Alessandra

3

Hello,

Here are some of the information on my DNA structure:

  • DNA structure is B-DNA with 12 bp and double stranded.
  • Average length is 3.8 nm long
  • The PBC structure is a triclinic with right angles at all plane system.
  • The water used is TIP3P with diethylene glycol as cosolvent. I used AMBER force field.
  • As of this moment, I increased my PBC to 15 nm since I received again a fatal error regarding inconsistent shifts over periodic boundaries during the MD simulation proper when I used -d option of 10.0 nm (It happened at the end of the run- last 100 frames of my run)
  • The molecule type of 758 atoms is the DNA structure itself.

I only decoupled the van der Waals forces on my free energy calculation since I need to get the solvation energy of the DNA. I can increase the size of the PBC, but I am doing my thesis remotely so it is hard to get my results on the laboratory due to travel restrictions (the data generated is very large- estimated 18 GB for one value of lambda alone)

I do hope that the information that I’ve provided will help you all visualize my problem during the setup. I am grateful for your help on my thesis.

Thank you and I do hope for your safety!

Sincerely,
Alexis

4

Free energy of solvation requires transformation of both vdW and electrostatic terms; note that my tutorial (which many people follow) explicitly does not compute free energy of solvation, only one contribution to it. I think this confuses a lot of people.

Please note that in general, trying to apply an alchemical transformation to a biomolecule is unlikely to be successful, for both algorithmic stability reasons (as you’re encountering) and for convergence. You’re never likely to get a converged value. For free energy of biomolecular solvation, MM/PBSA is the generally utilized approach.

5

I see. Even that I increase the size of the PBC, I still encountered the same problem. I will try to read the MM/PBSA and use it for my free energy calculations.

Thank you for your help. I really appreciate it.