• ## Outreach

From: Timothy Isgro (timisgro_at_ks.uiuc.edu)
Date: Tue Mar 07 2006 - 20:02:29 CST

Hi Ana,

Thanks for the suggestions. They have been implemented in a new
version of the NAMD tutorial, which should be available soon. The
new tutorial is also available in a Windows version as well as a Unix/
Mac version. The new tutorial uses tcl scripting to perform analysis
of simulations, so all analysis can be done inside of VMD. The Unix/
Mac version still uses xmgrace for plotting data, and the Windows
version uses Microsoft Excel.

> 1) It would be very useful if you could provide a model
configuration file, for example taking the ubiquitin in a water box
or in a
> water sphere examples, of restart analyses. Since restart analyses
configuration files are very different from the initial files, it
> would save new users a lot of time if you could provide a
commented template for a restart file.

The file namd-tutorial-files/common/sample.conf is a template for
restarting simulations. A lengthy info box has now been included in
the tutorial to explain this to the user.

> 2) Explaining what PMEGridSizeX ( or PMEGridSizeY or PMEGridSizeZ)
is would benefit from a more detailed text and also from
> an example. Something like this is what I had in mind:
>
> For PMEGridSizeX, one should use a number as close as possible
(can be a bit bigger or smaller) to the number on
> cellBasisVector1, but one that can also be factorized in 2's, 3's
or 5's (the smaller the better). This will great a grid in which each
> grid point is 1 angstrom apart from the next, and will make the
Fast Fourier Transforms that exist in this part of the code much
> faster. A similar thing should be done for PMEGridSizeY and
PMEGridSizeZ. Here's an example: if the cellBasisVector1 is
> (63,0,0) then a good PMEGridSizeX would be 64 because 64=26. 60 is
less good because 60=2*2*3*5.

This has been changed. The new text now reads:
PMEGridSizeX: sets the size of the PME grid in the direction along
cellBasisVector1 (not the x direction as implied). It is the number
of grid points in the system in the direction specified by
cellBasisVector1. Since the grid will replicate the charge
distribution in your system, PMEGridSizeX should be chosen to be
large enough so that the grid spacing accurately reproduces your
charge distribution. However, it should not be so large that it will
slow down your simulation to provide unnecessary precision.
Typically, a grid size slightly less than 1 A is a good choice to
reproduce charge distribution in biological systems, where the
closest atoms have a bond separation on the order of 1 A.
Furthermore, for speed in computing Fast Fourier Transforms,
PMEGridSizeX should be chosen so that it can be factorized by 2, 3,
or 5. If your cellBasisVector1 = (60, 0, 0), a good choice for
PMEGridSizeX might be 64, since 60 A / 64 = 0.9375 A and 64=2^6.

Hope this helps,
Tim

**