GROMACS version: 2022
Hello everyone,
I would like to ask if anyone has experience calculating the contact autocorrelation function (CAF) in GROMACS using gmx hbond -contact, which is commonly used to estimate residence times.
I observe reasonable results when calculating the CAF between water molecules inside a zeolite. However, when I calculate the same function between water molecules and a fixed framework atom (e.g. oxygen atoms in the zeolite, which are essentially immobile), the results become unphysical. In particular, the C(t) curve drifts toward negative values (even for larger simulation times), which is inconsistent with the expected behavior and with my other results.
Has anyone encountered a similar issue when using gmx hbond -contact for mobile–fixed pairs? Is there any known limitation or special consideration for such cases?
The command used was:
gmx hbond -f nvt_10000ps_10000.trr -s nvt_10000ps_10000.tpr -ac residence_acf.xvg -r 0.35 -n All_Indexes.ndx -contact
The autocorrelation plot is attached here as an example
.
Thank you in advance