[gmx-users] g_rdf

[Yulian Gavrilov]

Dear Gromacs users,

I have a problem with g_rdf.
I try to calculate the rdf function for two groups:
1. Oxygen atom of one water molecule;
2. Oxygen atoms of all other water molecules in the box;

When I use different number of frames from the trajectory I get very
different rdf plots.
First peak (round  0.26 nm) varies from 3.9 to 18 g(r).
If I am not mistaken, this number related to the coordination number of
water and should be found 3.8-4 in any case.
Why there is such strong difference?
Is there a problem with time averaging?

I get the same problem if initially prepare the short versions of the
trajectory for g_rdf input

gmx rdf, VERSION 5.0.4

g_rdf -f traj.xtc -s md.tpr -o rdf_0to1000frame.xvg -n index.ndx -b 0 -e 100
g_rdf -f traj.xtc -s md.tpr -o rdf_0to1000frame.xvg -n index.ndx -b 0 -e
1000
g_rdf -f traj.xtc -s md.tpr -o rdf_0to1000frame.xvg -n index.ndx -b 0 -e
5000
g_rdf -f traj.xtc -s md.tpr -o rdf_0to1000frame.xvg -n index.ndx -b 0 -e
10000

png plot is attached

Thank you
-- 

Sincerely,

Yulian Gavrilov