From: Norman Geist (norman.geist_at_uni-greifswald.de)
Date: Wed Apr 16 2014 - 07:49:47 CDT
I guess the major problem wasn’t the charges, but that single hydrogens „ions“ are just too fast so will violate the patch margins within stepspercycle permanently ;)
Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im Auftrag von Vlastimil Zíma
Gesendet: Mittwoch, 16. April 2014 12:29
An: Namd Mailing List
Betreff: namd-l: Re: Persistent margin errors
Resolution: Finally I found where the bug was - the system was corrupt. When I removed extra water from my system I did it on per atom (not per molecule) basis in several cases, which left H and OH chunks of water molecules. These molecules hadn't integer charges with lead to non-neutral system. This was most likely the reason which manifested as margin errors.
Since I fixed this I haven't encountered these errors.
2014-02-26 13:38 GMT+01:00 Vlastimil Zíma <zima_at_karlov.mff.cuni.cz>:
I lately get stuck with my simulation on margin errors which repeats in my simulation.
I have a system of transmembrane protein in a lipid bilayer in water much like the one from NAMD membrane tutorial. I did fairly similar set of simulations as in tutorial, but after 0,5 ns equilibration with protein constrained and 3 ns of equilibration I still can not make a 1 ns simulation without margin error, not even with margin 10, which is in my experience pretty high, I usually get along with margin 2-3.
I though of useConstantArea option, but according to the XST files dimensions have not converged. On the other hand, system volume is pretty stable.
I tried to run the simulation with output every step, but these did not produce the margin error.
All configuration files and the last output are available here: https://gist.github.com/vzima/9228435
Not to be confused, files named *_run are still an equilibration.
Is it possible that there is a bug in pressure control which prefers extension in Z dimension at cost of reduction in X and Y dimensions?
Thanks for any suggestions
--- Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirus Schutz ist aktiv. http://www.avast.com
This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:21 CST