Tianxiao Yang wrote:
>
> Dear Ambers,
>
> Could you please tell me the best way to calculate the residence time of
> water molecules in the hydration shell? Many thanks in advance.
>
> Tianxiao Yang
> School of Engineering,
> The University of Tokyo
Detail analysis of hydration sites with evaluation of occupancies and
residence times were done in our group for several types of nucleic acid
molecules. We suggest the following procedure: At first, minimal distances
between a selected atom forming hydration site and all solvent molecules
are calculated during the trajectory using the carnal option DISTRIBUTION
MIN (hydration sites can be easily identified by hydration density maps
calculated by rdparm/ptraj and vizualized by MidasPlus). The closest
solvent molecules are filtered using a distance criterion (see examples in
Amber manual, section carnal). Then the distances between these selected
water molecules and hydration site forming atoms are calculated and
plotted. Thus we trace individual water molecules.
Note that while residency times of water molecules are usually on a scale
of 50-300 ps, anomalous sites with long-residency water molecules can be
found occasionally and these can play very important structural roles.
The longest residency time has been reported in our recent paper Spackova
et al, J. Am. Chem. Soc. 2000, 122, 7564-7572. In this simulation of
d(GCGAAAGC)2 duplex with sheared G.A mismatches and adenine zipper, we
found a critical hydration site stabilizing a local conformational
variation at the zipper-mismatch junction. The site has 100% occupancy,
and during 9ns trajectory, only four water-exchange events occured, with
individual residency times 0.5 to 5.1 ns.
Note also, that to analyze hydration properly requires to have LONG
simulations expanded beyond 5-10ns. I would in general suggest to skip
data from the first ns of production, as the system and its solvent are
still under a substantial equilibration.
Nada Spackova
-----------------------------------------
Nada Spackova
e-mail:spackova_at_ibp.cz
National Centre for Biomolecular Research
Institute of Biophysics
Academy of Sciences of the Czech Rep.
Kralovopolska 135
612 65 BRNO
Czech Republic
-----------------------------------------
Received on Thu May 17 2001 - 06:24:07 PDT
