NAMD Implicit GB Parameters

From: Samuel Bowerman (sbowerma+namd-l_at_hawk.iit.edu)
Date: Mon Feb 01 2016 - 12:41:03 CST

Hello NAMD community,

I am was hoping somebody could provide insight into how NAMD defines some
parameters for GB simulations. I see from the user manual that atomic
radii are taken from Bondi (1964). How exactly does NAMD apply these
corrections? Is it a direct definition from some table of values, or does
it apply a subraction/addition based upon atom type? I ask this because I
am wary of "double-counting" radii corrections when comparing the results
of NAMD implicit solvent simulations against other ones I have conducted in
pmemd.cuda.

More exactly, to do the simulations in Amber, I had to alter the radii
terms in the .prmtop file to mbondi2 (Onufriev, Bashford, and Case, *Proteins
*2004). These parameters are supposedly the best for use with the NAMD
default GB values. If I load this .prmtop into NAMD for a simulation, is
it "double-dipping" into radii corrections by applying a Bondi correction
to the mbondi2 terms, or are they explicitly Bondi radii?

Take care,
Samuel Bowerman
Ph.D. Candidate
Illinois Institute of Technology
Chicago, IL 60616

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:21:48 CST