[gmx-users] Re: protein unstable for parallel job while stable for serial one

Akansha Saxena asaxena17 at yahoo.com
Tue Sep 19 16:58:27 CEST 2006

Hi Tsjerk,

>Hi Akansha,

>Either you need to simulate both systems until you
>have full
>convergence before you can make conclusions regarding
>the (dis)similar behaviour 

Well I can simulate it until convergence but that on a
single processor for my system will take several
months of computer time.

>or you have to test consistently, running replicate
>simulations both on one processor and on 16.

This is what I was doing. I was running exactly
identical simulations on 1 processor and on 16
By identical i mean - same starting structure,
velocities taken from the same *.trr file. The only
difference was the number of nodes for the production

>That being said, it is very well possible that even
>between two
>simulations of the same system under completely
>identical conditions,
>except for the starting velocities and possibly the
>exact starting
>structure, you will get two completely different
>results. If you're
>protein can undergo a conformational change it's
>basically a matter of
>chance whether you observe it or not.

But I give the same velocities and use exactly same
starting structure for both simulations. Basically I
use the same files for both cases. Only difference
lying in the number of processors. 
I would think that with same intial conditions the
calculations should be identical for both cases.   

>Of course, I don't say it is not related to your
>simulation being
>distributed over multiple processors. It may be, but
>in order to
>exclude other effects you have to repeat simulations
>and/or work
>through the code very carefully. Which version of
>Gromacs did you use?

I am using Gromacs3.3.
It would very helpful if you can please suggest what
care should I take in order to exclude other effects
when I repeat the simulation.



>Hope it helps,
>On 9/18/06, Akansha Saxena <asaxena17 at yahoo.com>
>> Hello
>> Has anybody seen a protein becoming unstable on
>> parallel nodes while remaining stable on single
>> I am running a NPT molecular dynamics simulation.
>>The system contains a protein with 141 residues
>> by water making the total to 32714 atoms. I ran
>> simulation at two places for 8ns. One as a serial
>> and the other as a parallel job on 16 nodes.
>> The protein remains stable all through for the
>> job but for parallel job the protein  opens up
>> 3ns of simulation and is never stable after that.
>> All other conditions are exactly the same. Any help
>> would be highly appreciated.
>> Akansha
>> Akansha Saxena
>> Graduate Student
>> Department of Biomedical Engineering
>> Washington University in St Louis
>> gmx-users mailing list    gmx-users at gromacs.org
>> http://www.gromacs.org/mailman/listinfo/gmx-users
>> Please don't post (un)subscribe requests to the
>>list. Use the
>> www interface or send it to gmx-users-request at
>> Can't post? Read

>Tsjerk A. Wassenaar, Ph.D.
>Groningen Biomolecular Sciences and Biotechnology
>Institute (GBB)
>Dept. of Biophysical Chemistry
>University of Groningen
>Nijenborgh 4
>9747AG Groningen, The Netherlands
>+31 50 363 4336

Akansha Saxena
Graduate Student
Department of Biomedical Engineering
Washington University in St Louis

Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 

More information about the gromacs.org_gmx-users mailing list