RMSD shoots up

GROMACS version: 2020.2
GROMACS modification: Yes/No
Here post your question

Dear all,
I am running a simulation of a complex molecule. it has two chain. When I calculated the rmsd and found that rmsd value shoots up at 16ns and 25ns. In corresponding PDBs two chains are far apart, out of the box. Is it artifacts? if the complex broke at 16ns why it re associates ? How can I fix it? Thank you

This is a periodicity artifact. Recenter with trjconv.

Dear Dr Jalemkul,
Following your advice, I combined the xtc file using “gmx trjcat -f step5_*.xtc -o combined.xtc -settime”
then i tried to remove PBC using " gmx trjconv -s step4.0_minimization.tpr -f combined.xtc -o md_0_1_noPBC.xtc -pbc mol -center"
Then i calculated the rmsd using “gmx rms -s step4.0_minimization.tpr -f md_0_1_noPBC.xtc -o rmsd1.xvg -tu ns”
but still its not centered. the two chains are far apart and rmsd is absurd for these frames. Am I doing anything wrong?
Thank you


For multiple chains, center on one of them, but you will have to do other steps first as suggested in this link.

Thank you Dr Jalemkul. As I am new to Gromacs. I really couldn’t follow the steps given in the link. Is it possible to get set of commnads. Sorry for the trouble.
Best regards

You will learn lots by reading the trjconv help information, which describes everything mentioned in that link :)

Dear Dr Jalemkul,

I mange to fix pbc issue. Thanks for your help. I got another question. what is criterion to decide the simulation has saturated? Is the simulation below saturated already at 40ns? as the rmsd plot is flat though i still see small bumps

Are you talking about convergence? RMSD doesn’t tell you that. It’s a degenerate metric. You need to examine all relevant structural properties of interest and they should be statistically invariant over time.

1 Like