I need to estimate the mobility of some atoms in the molecule within a molecular dynamics simulation and compare the values for different simulations. I use trjconv -fit rot+trans to fit C-alpha atoms in a trajectory and aim to analyze some atoms in different residues (for example, C-beta). What is the correct way?
- dynamic RMSD (gmx rms -prev)
- mean square displacement (gmx msd)
- RMSF (gmx rmsf)
Anyway, is it correct for a set of atoms (by making a ndx group for all C-beta atoms) or I should calculate values for separate atoms/residues and then compare average values (or individual values?)?
Hi Pavel! I wouldn't use just one way to estimate the mobility but try to combine several ways. Since you mention C-alpha, I assume that you are dealing with proteins. One of the way of investigating the motion of proteins is to see the changes in secondary structure in time as well as changes in phi/psi dihedrals in time. The second way is to calculate van Hove's time dependent correlation function and self-intermediate scattering functions. The third way is to compute rotational and re-orientational time correlation functions. Those ways of investigation can be combined together with what you have mentioned in order to provide the information about the mobility of atoms and dynamics in the investigated systems. All those methods would require you to create index files. How the data should look like in the index files you may see in GROMACS manual for particular routines. Hope I could help you.
- Badges
- Science topic