charges in psf and in namd (or vmd)

Date: Wed Jun 02 2010 - 11:12:07 CDT

Dear NAMD developers:

There is a issue with the way of how NAMD or VMD gets the charge of any
particular atom or system. As far as I understand the psf file defines the
partial charges but there is always a slight difference between the psf
partial charges and the one that is computed in VMD or NAMD. For example in
the attached topology file the Toluene total charge is zero, but if you run
this molecule in NAMD ,in the output, is stated that the toluene molecule has
a total charge of -3.725290298461914e-9 fact that is also obtained by using
VMD. To confirm this I checked the charges of each atom in VMD and also there
is an small difference, for example atom C7 in the topology file (psf) is
-0.06 but in VMD is -0.06499999761581421, it the same with the rest of the

I know that for small systems this is irrelevant because the charge is almost
equal but I as you increase the size of your system this charge difference
might be important specifically for the electroneutrality of the system. In
particular I am runnig a 1.7 million atom system of toluene and at this level
the difference between the psf total charge and the charge computed by NAMD is
very close to one. Therefore my questions are two:

Why there are this small differences?
Should I be worried about this or the actual charges are the ones in the psf

Jose Antonio Garate

PhD, Molecular Simulation Group
Engineering Building, University College Dublin.


This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:54:11 CST