From: P.-L. Chau (pc104_at_pasteur.fr)
Date: Thu Jul 04 2013 - 16:43:32 CDT
I write in relation to the NAMD calculation of pressure profiles, using
equation 2.5 of a paper by Lindahl and Edholm in 2000.
The pressure depends on two terms. The 1st term depends on the mass and
velocities of the particles, and the 2nd term depends on the position of,
and forces on, the particles. NAMD calculates the 2nd term correctly, but
when it calculates the 1st term, it does not use the velocities of the
particles (from a velDCD file, example) but assigns them randomly. This
situation has already been addressed previously by Grace Brannigan in
August 2012.
I am thinking of overcoming this by calculating the 1st term oneself. If
one outputs a velDCD file in addition to a DCD file, one can access the
velocities to calculate the 1st term using a home-made programme. The 2nd
term can then be calculated using a very low temperature (so that the
velocities are nearly zero), and it has to be done twice, once for the
Ewald component and the other time for the non-Ewald component. The three
sets of results are added together to yield the pressure profile. Would
this suggestion work?
Thank you.
P-L Chau
email: pc104_at_pasteur.fr
Bioinformatique Structurale
CNRS URA 2185
Institut Pasteur
75724 Paris
France
This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:23:26 CST