[AMBER] Normalization of radial distribution function

From: Szymon Żaczek <szymonzaczek.gmail.com>
Date: Sun, 3 Dec 2017 19:38:08 +0100

Dear Amber users and developers,

I am trying to get an insight into how interactions of aromatic
residues evolve throughout simulation of my system by studying radial
distribution functions. My system consists of 6 identical protein
chains with size of 104 AA each and 30400 molecules of water - overall
there are 100674 atoms. The whole dynamics took 500 ns. Both density
and RMSD values behave normally. What I want to do is to compare RDF's
of carbon atoms in PHE and TRP around carbon atoms in PHE and TRP
from the first 20 ns and last 20 ns of simulation.

>From what I understand about RDFs, if they are properly normalized,
their values should be very close to 1 when the distance is
approaching infinity. However, I failed in achieving this. I tried
calculating RDFs using both volume and density options, also with
specifying different numbers for density, however none can get me
anywhere near achieving value close to 1 - I always end up with
values close 0 in distance greater than 80 A.

Could you kindly advise me what is the the best way to achieve
satisfactory normalization? Otherwise, if I got something wrong - if
at such distances and for a densely packed system like mine with big
protein chains it is normal to achieve such values, please also inform

Below there is one of the scripts I tried for calculating RDFs:

parm ../../box_monomer.prmtop
trajin ../production.15.mdcrd 2000 last
radial averaged_PHE_PHE_production_15.rdf 0.1 150
:PHE.CE1,CD1,CG,CD2,CE2,CZ :PHE.CE1,CD1,CG,CD2,CE2,CZ volume
radial averaged_TRP_TRP_production_15.rdf 0.1 150
:TRP.CE2,CD2,CE3,CZ3,CH2,CZ2 :TRP.CE2,CD2,CE3,CZ3,CH2,CZ2 volume

If any of the files are needed to properly troubleshoot the problem,
please, inform me.

Kind regards,
Szymon Zaczek

AMBER mailing list
Received on Sun Dec 03 2017 - 11:00:02 PST
Custom Search