Trajectory preprocessing in peptide self-assembly simulations

User discussions
1

GROMACS version:
GROMACS modification: No

Dear all,

I performed molecular dynamics simulations of peptide self-assembly using GROMACS. My system consists of 150 antimicrobial peptides (13 amino acids each) simulated in explicit water with periodic boundary conditions.

For a standard protein simulation, I usually preprocess the trajectory using:

gmx trjconv -pbc mol -center -ur compact

However, in a peptide self-assembly simulation there are many independent peptide molecules that gradually form aggregates, and I am concerned that some trajectory manipulations (especially centering or clustering) might artificially alter the relative positions of the aggregates.

What would be the recommended preprocessing workflow for this type of simulation before quantitative analyses such as cluster analysis, radius of gyration, intermolecular contacts, hydrogen bonds, and SASA?

For example:

  • Should I first apply -pbc whole?

  • Is -pbc mol -center -ur compact still appropriate for this type of system?

  • Should -pbc cluster be avoided unless it is used only for visualization?

  • Is it preferable to keep the raw trajectory unchanged and simply extract the peptide coordinates?

I would appreciate any recommendations or best practices for preprocessing trajectories of multi-peptide aggregation simulations prior to analysis.

Thank you very much.