Re: Na ions got into the catalytic site during Production Run

From: Marc Q. Ma (qma_at_oak.njit.edu)
Date: Mon Nov 21 2005 - 09:29:51 CST

Jasmine,

Since your Na+ ions are just counterions, they should not be in the
catalytic site, or anywhere near it. If indeed these counterions come
closer to or even enter these "not allowed" regions, your production
runs would be seriously flawed.

If "autoionize" does not place these counterions far away from your
catalytic site, you can do it manually. i.e., pick some locations of
water molecules that are "behind" the catalytic site of the protein,
replace them with Na, and regenerate the PSF and PDB files. when you do
these, you should make sure that the distribution of the ions are
distant from each other, do not make them too artificial -- like a Na
cluster ...

The immediate resulting system can not be used for production run. You
have to run energy minimization (maybe fixing protein first), and
equilibration.

Marc
On Nov 20, 2005, at 6:34 PM, Ching Wong wrote:

> Hi. I am just a beginner. Please help. :)
>
> My protein is negatively charged and I added eleven Na ions to zero out
> the charges prior to energy minimization. Then I realize 3 Na ions got
> into the catalytic site during the produciton run. I checked back with
> the dcd files during the earlier energy minimization runs, one Na ion
> already got in there. Then I restart the whole thing by fixing the
> positions of all the Na ions (total 11) during the energy minimization
> stage, then I proceed on production run. Again one Na ions got in the
> catalytic site. My questions are
>
> 1.) is that okay to fix the Na ions during energy minimization?
> 2.) the Na ion got into the catalytic site again, should that be taken
> into account when interpreting the rmsf result?
> 3.) My bigger problem is: the production run (with Na fixed during
> energy minimization) actually crashed. (i got enough frames to realize
> Na ion already got into the catalytic site). I wasn't able to use the
> vel or xsc files from the previous Heating and equilibrium run. (I
> could use that for my production run when I didn't fix the Na ions
> during prep) The error shown in the log file is :
> FATAL ERROR: Periodic cell has become too small for original patch
> grid!
> Possible solutions are to restart from a recent checkpoint,
> increase margin, or disable useFlexibleCell for liquid simulation.
>
> My conf file:
> My cellBasisVectors are {68, 68, 82}
> PMEGridSize are {80,80,80}
>
> dcd minmax
> frame 0 {33, 33, 40}
> frame 489 {36.7, 36.7, 32.9}
>
> 4) My other doubt is how to set the cellBasisVector. I realize the cell
> got bigger after waterMD (600K). It got smaller during production run.
> So, how to set up the right CellBasisVectors? Do you think if that's
> the reason my Production run crashed?
>
> Thanks in advance for helping to solve these problems.
>
> Sincerely,
> Jasmine
>
> -------
> Jasmine
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:41:23 CST