RE: Calculation of CHARMM Parameters from GAMESS Output.

From: JC Gumbart (
Date: Mon Feb 05 2007 - 10:15:40 CST

If your frequency is in wavenumbers, shouldn't there be a factor of c^2 in
your equation as well?


Anyway, the conversion is 1 kcal/mol A^2 = 69.52 pN/A. You have to throw
Avogadro's number (6.022*10^23 atoms/mole) in there to make it work.



From: [] On Behalf
Of Klon, Anthony
Sent: Monday, February 05, 2007 9:27 AM
Subject: namd-l: Calculation of CHARMM Parameters from GAMESS Output.


I have a rather silly question. I've been trying to calculate CHARMM
parameters for a small molecule that isn't defined in any of the topology /
parameter files. Although CHARMM parameters can be built up in an additive
fashion, there are still some groups and hence some atom bond lengths /
angle terms which are not defined.

Geometry optimization of these fragments proceeded smoothly in GAMESS, and
once a stable point was obtained, the full Hessian matrix was calculated and
the frequencies were extracted.

Here's the problem:

To calculate the bond force constant from a frequency, it seems as though it
should be a straightforward exercise:

k = 4 * pi ^ 2 * u * v ^ 2

where u is the reduced mass, and v is the frequency (actually, GAMESS
outputs the wavenumber in cm ^ -1).

The first problem is that I can't seem to get my units to match up. GAMESS
reports frequencies in cm ^ -1, and reduced masses in amu (atomic mass
units). The constants in the CHARMM parameter files are in units of kcal /
mol / Ang. ^ 2.

The wavenumbers can obviously be converted into angstroms, and the amu's
into units of kcal, but this still leaves something out (namely that there
are no moles in the calculated result!).

I know that other NAMD users must have dealt with the same problem. I'm
probably missing something trivial, but don't know what it is.



Scanned by IBM Email Security Management Services powered by MessageLabs.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:21 CST