GROMACS version: 2024.4
GROMACS modification: No
Hi all,
I’m running a simulation involving a 20 base pair DNA molecule in a triclinic box using semi-isotropic pressure coupling with the c-rescale barostat. To prevent end fraying and maintain DNA stability, I applied distance restraints on specific base pairs. Since the DNA needs to remain structurally intact, I used pdb2gmx -merge all to treat it as a single molecule.
This setup is part of a planned peptide–DNA interaction study, where it’s critical that the DNA ends do not melt; otherwise, the peptide may bind to an artificially frayed region, leading to misleading interaction results.
After running the simulation, I noticed the following during post-processing:
- In npt.gro, the DNA stays inside the box, but when I visualize it together with md.xtc, it appears to have shifted closer to the boundary and slightly crosses the box edge.
- Using
-pbc nojumpin trajectory processing causes the DNA to break visually, likely becausenojumpdoesn’t preserve molecular connectivity. - Using
-pbc mol -center -ur tricgives a cleaner structure, but the DNA still appears partially outside the unit cell. - With
-pbc wholefollowed by-center -ur tric, the DNA is whole and centered, but some atoms are still outside the visible triclinic box, which makes interpretation tricky.
I’ve attached images showing the behavior.
What I’d love expert insight on:
-
Is this minor visual box-crossing expected behavior for a rigid molecule like DNA under semi-isotropic pressure in a triclinic box?
-
Are there recommended best practices for handling DNA as a single molecule with restraints and semi-isotropic coupling?
Thanks in advance for your insights and suggestions.
