Rotational Autocorrelation calculation with Center(s) of Mass in ligand

GROMACS version: 2020.4
GROMACS modification: No

Hi all,

I want to use “gmx rotacf” lately to calculate the rotational autocorrelation of a nucleotide and I wonder if I can use CoM from different parts of the ligand, e.g. phosphate, ribose and base, to calculate the rotational autocorrelation, instead of atom triplets in rotacf?

I thought about calculating the cross product of vectors between CoM as output from “gmx trajectory” and then do autocorrelation in “gmx analyze” but It seems gmx analyze does not support vector autocorrelation calculation. As I check the source code for rotacf (as in src/gromacs/gmxana/gmx_rotacf.cpp), it is actually doing vector autocorrelation calculation, but I am clueless on how to pull up the related functions.

Thanks in advance!

Oh I later figured out how to do this in a simpler way in case anyone is interested.

I first calculate the CoMs and output their coordinates in xtc with gmx traj. Then I can use gmx rotacf to calculate the ACF.