Re: Why do different versions of NAMD deserve different simulation results?

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Mon Dec 20 2010 - 06:48:27 CST

yuhui,

[...]

> At first, I thought there might be some atoms which were too close to
> each other, leading to extreme values of vdW energy. However, I checked the
> PDB file by writing some VMD codes, and found no such case.

have you also checked for close contacts due to periodic boundary
conditions, e.g. by providing unreasonable simulation cell dimensions?

> The more incredible finding is that when I run the same namd script
> with the same PDB and PSF file on the workstation, where
> NAMD_2.7b1_Linux_x86_64 is installed, the above artifact never happens!
>
>
> LDB: ============== END OF LOAD BALANCING =============== 518.576
>
> Info: Initial time: 1 CPUs 5.16644 s/step 59.7968 days/ns 148.165 MB memory
> TIMING: 100 CPU: 514.58, 5.0948/step Wall: 521.851, 5.16675/step, 0 hours
> remaining, 148.165436 MB of memory in use.
> ETITLE: TS BOND ANGLE DIHED
> IMPRP ELECT VDW BOUNDARY
> MISC KINETIC TOTAL TEMP
> POTENTIAL TOTAL3 TEMPAVG PRESSURE
> GPRESSURE VOLUME PRESSAVG GPRESSAVG
>
> ENERGY: 100 143463.3277 12983.6697 0.0000
> 0.0000 -2525607.1884 832652.3203 0.0000
> 0.0000 0.0000 -1536507.8707 0.0000 -1536507.8707
> -1536507.8707 0.0000 9770920.1156 9774414.2593
> 1208309.6819 9770920.1156 9774414.2593

well, that doesn't look much better. you have an incredibly high pressure
and ...

> However, it reports "bad global exclusion" when I run the following
> simulation of raising temperature on the workstation.

.. this confirms that your system is still in a very bad shape.

cheers,
    axel.

> Can anyone give some explanations?
>
> Thanks and Best wishes!
>
> Yuhui
>
>
>
>
> --
> Ph. D Yuhui He (何毓辉)
> Institute of Scientific and Industrial Research, Osaka University;
> Mihogaoka 8-1, Ibaraki, Osaka 567-0047. JAPAN
>
>

-- 
Dr. Axel Kohlmeyer
akohlmey_at_gmail.com  http://goo.gl/1wk0
Institute for Computational Molecular Science
Temple University, Philadelphia PA, USA.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:54:53 CST