[gmx-users] Contact autocorrelation function in Gromacs
ebadi.mahsa at gmail.com
Wed Apr 10 16:10:20 CEST 2019
I want to calculate the contact duration of ions with special atoms in
polymer chains within 3 Å of the ion.
I found gmx hbond as a good option for this analysis, but I have some
questions regarding this calculation. By the way, the version of gromacs is
1. I used the following command:
gmx hbond -s topol.tpr -f md.xtc -n index_contact.ndx -contact -r 0.3 -num
contact.xvg -g contact.log -ac contact_ac.xvg
in the gromacs manual it is not clear whether r or r2 should be used for
this analysis. I tried this as well:
gmx hbond -s topol.tpr -f md.xtc -n index_contact.ndx -contact -r2 0.3
-num contact.xvg -g contact.log -ac contact_ac.xvg
The final results were little bit different from the first one. Which one
is correct -r or -r2?
2. At the end of this calculation, I get the following warning:
WARNING: Correlation function is probably not long enough
because the standard deviation in the tail of C(t) > 0.001
I used the trajectory of 500 ns simulation time but still get this warning.
I also tried trajectory with saving every frame but still get the warning.
How can I solve this?
3. The tail of ACF goes to negative values, I found this link for the
Does this means that I should consider the second column of y values in the
contac_ac.xvg file instead of the first one?
More information about the gromacs.org_gmx-users