Re: PMEGridSize and parameters controlling electrostatic interactions.

From: Maxim Belkin (mbelkin_at_ks.uiuc.edu)
Date: Sun Jan 25 2015 - 21:20:58 CST

Yes, "pmeGridSpacing 1.0" is just a good practice, not a solution to “Low global exclusion count” problem. I think the problem could be caused by some bad settings. Is this an NPT or NVT? Dawid, before rebuilding the system, try increasing margin, let’s say, to 10 Å.
Also, you might want to minimize and equilibrate the system at your lowest temperature and then proceed with the heating-up procedure.

> On Jan 25, 2015, at 18:11, Jeff Comer <jeffcomer_at_gmail.com> wrote:
>
> Hi Dawid,
>
> "Low global exclusion count!" is usually a problem with the
> preparation of the system. Often it means that you've chosen the
> system size incorrectly or you have some kind of clash in the system
> that minimization wasn't able to work out. Maybe you have a alkyl
> chain stuck inside of a phenol ring somewhere. You can try looking for
> problems like this, or if that doesn't work, try rebuilding the
> system.
>
> pmeGridSpacing 1.0 means that NAMD picks the pmeGridSizes for you,
> which is a good idea, but probably won't fix the problem.
>
> Jeff
>
>
> –––––––––––––––––––––––––––––––––––———————
> Jeffrey Comer, PhD
> Assistant Professor
> Institute of Computational Comparative Medicine
> Nanotechnology Innovation Center of Kansas State
> Kansas State University
> Office: P-213 Mosier Hall
> Phone: 785-532-6311
>
>
> On Sat, Jan 24, 2015 at 1:48 PM, Dawid das <addiw7_at_googlemail.com> wrote:
>> Dear NAMD experts,
>>
>> I was doing MD simulation on a small system and I used following parameters:
>>
>> Pme on
>> PmeGridSizeX 48
>> PmeGridSizeY 48
>> PmeGridSizeZ 96
>>
>> switching on
>> switchDist 10
>> cutoff 12
>> pairlistdist 14
>> margin 3
>>
>> Everything went fine, until I made my system bigger, so I had to change PME
>> grid sizes to:
>>
>> PmeGridSizeX 96
>> PmeGridSizeY 96
>> PmeGridSizeZ 244
>>
>> Now when I try to heat up my system I get following error at 116 K:
>>
>> FATAL ERROR: Low global exclusion count! (1086603 vs 1086604) System
>> unstable or pairlistdist or cutoff too small.
>>
>> So my question is if there is any relationship between PmeGridSize and
>> electrostatic parameters values in a sense that I should change the latter
>> in this case?
>> Or perhaps I shall not increase PmeGridSize, but PMEGridSpacing instead?
>>
>> Best wishes,
>>
>> Dawid Grabarek

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:20:50 CST