AMBER Archive (2009)

Subject: Re: [AMBER] RE: question

From: Marek Maly (marek.maly_at_ujep.cz)
Date: Thu Nov 05 2009 - 07:28:03 CST

Hi Thomas,

thanks a lot for your advice regarding to ntt value ! I hope that
it helps to Pavel to solve the problem.
Anyway there is no warning in Amber10 manual regarding to
ibelly/ntt=3 issue which you mentioned. There is just reported
that igb>0 is incompatible with this constraint, so it was hard
for us to find out cause of "moving problem" and his relation
to Langevin thermostat. So let's wait for Pavel's verification ...

Of course that Pavel tried to use standard harmonic restraint approach but
there is problem
that cca restraint_wt=7000 was necessary to prevent surface from
"explosion" and this high value of restraint_wt
and so also possible high forces involved directly into simulation is also
not the optimal solution...

Such a big restraint_wt value is necessary to keep the our diamond
structure stable mainly because
we used as simplest model as is possible to simplify loading of the
surface into Leap, it means
that we have no internal bonds which can keep the crystal structure
compact. Also the right VdW
parameters for carbon atoms in diamond could help a little bit, but the
main problems are of course
missing bonds. So I think, that for this simple surface model "ibelly"
approach is the proper one.

There could be just one big disadvantage comparing to standard spring
restraints.
As was recently reported by Ross, PMEMD is also capable of calculations
with restrained systems
but just using old restraint syntax. Question is if this is true also for
absolut freezing with
"ibelly" approach.

Anyway I think, that the total atom freezing ( ibelly/bellymask ) should
not be eliminated from
the Amber (in the next versions) even if the main usage could lie just for
such a simplifications
of solid components as we did. I think that we are not the first and of
course not the last one who want
to use Amber for simulation of biomolecule + solid surface.

Thanks again !

     Best,

        Marek

Dne Thu, 05 Nov 2009 09:12:47 +0100 <steinbrt_at_rci.rutgers.edu> napsal/-a:

> Hi,
>
>> How wrote Marek, we have problem with freeze atoms. I use ibelly for it
>> and after simulation, atoms aren't on same place as on the beginning of
>
> IIRC, ibelly appears to be on the way to depreciation and people are
> encouraged to use harmonic restraints instead. Even weak restraints will
> keep your atoms almost as well fixed as ibelly would, but without
> artificially removing degrees of freedom.
>
>> &cntrl
> ...
>> ntt=3,
>> gamma_ln=2.0,
> ...
>
> This is were the problem lies in your run, I believe ntt=3 doesnt check
> for ibelly settings and still has virtual particles collide with the
> supposedly non-moving atoms. This makes them move a bit, consistent with
> your pictures.
>
> Use ntt=1 if you must use ibelly (but be aware that this may introduce
> problems as well), or switch to restraints.
>
> Kind Regards,
>
> Thomas
>
> Dr. Thomas Steinbrecher
> BioMaps Institute
> Rutgers University
> 610 Taylor Rd.
> Piscataway, NJ 08854
>
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
> __________ Informace od NOD32 4051 (20090504) __________
>
> Tato zprava byla proverena antivirovym systemem NOD32.
> http://www.nod32.cz
>
> 

-- 
Tato zpráva byla vytvořena převratným poštovním klientem Opery:  
http://www.opera.com/mail/

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber