Re: [AMBER] R: problem with pmemd.cuda

From: Pavan Giovanni <giovanni.pavan.supsi.ch>
Date: Fri, 11 Nov 2011 19:36:45 +0100

Dear Scott, dear David, dear all,

I am conducting several attempts following your
indications.
Probably it is still too soon to be sure 100% that the
problem is solved.
But it seems that Scott is right.
If I delete iwrap=1, the simulation goes from 1st to 2nd
ns without problems.
On the other hand, even if I substitute -ref RST with -ref
INPCRD the same problem occurrs, indicating evidently some
problem in the restart.
In the latter case, the simulation goes on only if I don't
split the entire simulation into 1 ns intervals.
I hope that setting iwrap=0 the problem is definitely
solved.
For the moment, many thanks to all of you for the
exceptional support!

best
g

On Thu, 10 Nov 2011 15:23:08 -0800
  Scott Le Grand <varelse2005.gmail.com> wrote:
> I think he has a case where the constraint coordinates
>are unwrapped but
> the restart files have been wrapped...
>
> And that'll blow things up... If the problem goes away
>with iwrap=0, case
> closed...
>
>
>
> On Thu, Nov 10, 2011 at 5:22 AM, case
><case.biomaps.rutgers.edu> wrote:
>
>> On Thu, Nov 10, 2011, Giovanni Pavan wrote:
>> >
>> > equil SAM
>> > &cntrl
>> > ntp=1,
>> > ntr=1,
>> > /
>> >
>> > What is really curious...is that it seems that the
>>problem occurs only at
>> > the restart
>>
>> This has the hallmarks of a problem with the restraint
>>coordinates: in the
>> restart run, what file are you using for the -ref flag?
>> If it is not the
>> same file as that for the -c flag, run an experiment to
>>see if the problem
>> goes away when -ref and -c point to the same file.
>>
>> This is just a guess, but it's simple to check.
>>
>> ....dac
>>
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Nov 11 2011 - 11:00:03 PST
Custom Search