From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Mon Jul 29 2013 - 10:31:22 CDT
On Mon, Jul 29, 2013 at 5:13 PM, Markus Dahlgren
<markus.dahlgren_at_yale.edu> wrote:
> Hi,
>
> Does a NAMD trajectory need to be wrapped/unwrapped prior to evaluation
> using NAMD Energy? I get different energies depending on whether the
how different?
trajectory data is stored, read and processed inside of VMD with
single precision floating point math and particularly for larger data
sets, this can result in significant deviations due to truncation
errors.
in a single precision floating point number you have 23 bits for the
mantissa. thus you can have the full resolution for coordinates where
their magnitude lies between 1.0 and 2.0. for 2.0 and 4.0 you have
half the resolution and thus lose one bit and similarly when you go to
absolute coordinate positions of 8, 16, 32, 64, and 128 angstrom. at
this point you lost 6 bits already and have 17 bits left and the
granularity of position data is about 7.5e-6 almost two orders of
magnitude larger than close to the origin.
axel.
> trajectory has been unwrapped prior to using NAMD Energy. Visually I can see
> that parts of the protein-ligand complex has crossed the boundary border and
> appears on the other side of the water box. Does the NAMD trajectory have
> only the central image coordinates? Is wrapping not necessary for NAMD
> Energy trajectory analysis when inputting NAMD trajectories? I have been
> using NAMD
> 2.8.
>
> -Markus
>
-- Dr. Axel Kohlmeyer akohlmey_at_gmail.com http://goo.gl/1wk0 International Centre for Theoretical Physics, Trieste. Italy.
This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:28 CST