[gmx-users] Regarding atom selection while using g_rdf

[soumadwip ghosh]

 Dear all,
                I am willing to calculate the rdf of certain cations
from different
group of molecules in the system. For example, I have got a DNA backbone
(having P, O1P and O2P atoms), DNA major groove (having all the atoms a
guanine and cytosine residue can have together) and a minor groove (having
all the atoms an adenine and thymine residue can have together). Now I want
to calculate g_rdf between Na/ backbone, Na/minor and Na/major.If I opt for
the -com option, it selects the centre of mass atom of the specific
groove/backbone group and while calculating PMF from g(r) I get a nice
curve as any of the PMF vs RC curve should look like. But as Dr Justin
suggested, I should not use the -com option in this case for residues with
specific interests. If I execute g_rdf of Na and minor/major groove/
backbone without using the -com option, I get some weird shaped PMF curves
which has a minima and it converges but it is of strange shape. I dont know
without -com how rdf is calculated between a residue like a minor groove of
a DNA (containing 114 atoms) and Na or other ions(22 in the system) since
it is atom pair correlation by definition . I dont really get how RDF is
calculated in my case. It will be highly appreciated if someone guides me a
bit since the present work of mine is based upon TD quantities calculation
from PMF and I have to ensure that the data obtained from g_rdf are
reliable.

Thanks and regards,
Soumadwip Ghosh,
Research scholar,
Indian Institute Of Technology Bombay,
India.