Re: NAMD Simulation error

From: Anton Arkhipov (anton_at_ks.uiuc.edu)
Date: Fri Nov 14 2008 - 15:10:27 CST

Anirban,

How about calculating the magnitude of the velocity of the atom that
NAMD says is moving too fast? The velocity vector has three
components, which NAMD prints out for you. You can use those to obtain
the magnitude of the velocity and see whether it is over or below the
limit, which is 1000.

Anyway, the problem is that at some point your atoms are moving too
fast. Without knowing what you are trying to simulate, and without
looking at the trajectory (which you should check carefully), it is
impossible to say why you are getting this error.

One problem is obvious from your letter, though. You are running your
residue-based CG simulation without explicit solvent, which is
completely unjustified. The residue-based CG method has been developed
to simulate coarse-grained systems in explicit, although coarse-
grained, water. So you need to solvate and ionize your system,
otherwise your simulation most probably does not make sense. It is
very possible that the error you are getting is due to the fact that
you don't have solvent in your simulations.

Keep in mind that there are additional potential restrictions on
applications of the residue-based CG. For example, the residue-based
CG force-field by itself will not maintain the tertiary structure of
proteins. One needs to add additional forces to maintain such
structures. Read carefully the following papers, and see references
there:

Assembly of lipids and proteins into lipoprotein particles. Amy Y.
Shih, Anton Arkhipov, Peter L. Freddolino, Stephen G. Sligar, and
Klaus Schulten. Journal of Physical Chemistry B, 111:11095-11104, 2007.
Disassembly of nanodiscs with cholate. Amy Y. Shih, Peter L.
Freddolino, Stephen G. Sligar, and Klaus Schulten. Nano Letters,
7:1692-1696, 2007.
Four-scale description of membrane sculpting by BAR domains. Anton
Arkhipov, Ying Yin, and Klaus Schulten. Biophysical Journal,
95:2806-2821, 2008.

Molecular models need to be tested: the case of a solar flares
discoidal HDL model. Amy Y. Shih, Stephen G. Sligar, and Klaus
Schulten. Biophysical Journal, 94:L87-L89, 2008.

Best,

Anton.

On 14 Nov 2008, at 02:36, Anirban Ghosh wrote:

>
>
>>
>> Hi ALL,
>>
>> I am running a simulation of a residue based CG model of 856 CG
>> beads in
>> NAMD. I am running it using implicit solvent condition without PBC
>> and
>> PME. After 5000 steps of minimization the system ran for 550000
>> steps and
>> gave a error:
>> ------------------------------------------------------------
>> FINISHED WRITING RESTART VELOCITIES
>> LDB: LOAD: AVG 0.110926 MAX 0.150218 MSGS: TOTAL 38 MAXC 7 MAXP
>> 7 None
>> LDB: LOAD: AVG 0.110926 MAX 0.112707 MSGS: TOTAL 38 MAXC 7 MAXP 7
>> Refine
>> ERROR: Atom 744 velocity is 684.102 -844.858 588.241 (limit is 1000)
>> ERROR: Atom 750 velocity is -555.261 769.693 -387.632 (limit is 1000)
>> ERROR: Atoms moving too fast; simulation has become unstable.
>> ERROR: Exiting prematurely.
>> ==========================================
>> WallClock: 929.760376 CPUTime: 643.129211 Memory: 7608 kB
>> -------------------------------------------------------------
>>
>> In the error the atom velocities did not cross 1000 (which is shown
>> to be
>> the limit) but still it terminated. I am getting the same error but
>> at a
>> very early stage if I use PBC and PME. I have tried increasing the
>> box
>> size also and now without PBC, but every time I am getting this
>> error. The
>> configuration script is attached. Please suggest a solution. Thanks
>>
>> Regards,
>>
>> --
>> Anirban Ghosh
>> Grade Based Engineer
>> Bioinformatics Team
>> Scientific & Engineering Computing Group
>> Centre for Development of Advanced Computing
>> Pune, India
>>
>>
>>
>
> --
> Anirban Ghosh
> Grade Based Engineer
> Bioinformatics Team
> Scientific & Engineering Computing Group
> Centre for Development of Advanced Computing
> Pune, India
>
>
> <CG.conf>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:21:29 CST