Changing topology/indices to match two systems

User discussions
1

GROMACS version: 2020.6
GROMACS modification: No

I have run a set of 6 simulations of a GPCR in a solvated bilayer, but I have a problem:

In the first 3 simulations I had a longer N-Terminus on my protein, which I removed for the latter 3 simulations as it was poorly modelled. I generated all the GROMACS inputs via CHARMM-GUI.

Now I’d like to prepare a meta-trajectory of all the simulations to align and perform cartesian PCA on, but I have extra residues in 3 of my simulations so there’s an issue. Is there some way I can remove the residues from 3 simulations and reorder the indices so that they match the other 3?

2

Technically, sure, almost every Gromacs trajectory processing/analysis tool asks you for a selection for the output - just create two index (.ndx) groups for the two systems that will have the same atoms selected, and pass your trajectories through gmx trjconv with them. No need to renumber atoms in .xtc, this format only keeps track of the total number of atoms.

1 Like
3

Thank you. I got it working via the following similar procedure:

  1. Create an index group removing the extra atoms from Protein in the extra-length receptor
  2. Create an index group removing the first 6 atoms from the normal length protein (to remove the N-terminal patch which isn’t present in the first set after truncation)
  3. Convert-tpr using the new index group and then trjconv in the same manner (perhaps the former is unnecessary)

After that it’s the usual rototranslational fit etc. for building a meta-trajectory for PCA.