Re: peridic boundary condition

From: Marc Q. Ma (qma_at_oak.njit.edu)
Date: Fri Mar 11 2005 - 14:00:37 CST

When you first build the system, you do not need a pbc. you can use
full electrostatics without using PME.

After you solvate your model, you want the water box to be larger than
the protein -- if you are working with a protein, say, 3AA or 5 AA
thick of water. This should prevent the protein getting out of the box.
Min can be done first on protein, then protein and water.

Your basis vectors has to be the same or just a little larger than the
water box to avoid bubbles or drop of densities.

Marc

> Then the problem is :
> (1) I have to anticipate what may be the model size for the last stage
> of
> simulation and accordingly i have to select cell basis vector.
> (2) If model size at the final stage of simulation changes considerably
> with respect to the initial structure then the cell basis vector fixed
> on
> that final size of model will not be oversized at the initial stages of
> simulation and lead to a wrong simulation result?
> (3) what will be the case if cell basis vector is fixed according to
> the
> model size at each stages of simulation?.....i mean what will give
> better
> result... keeping same cell basis vector or varying that at different
> stages of simulation according to model size.
> I am eagerly waiting for ur reponse.
> Thanks
> debashis
>> yes.
>>
>> Marc
>> On Mar 11, 2005, at 1:43 PM, debashis.sikdar_at_ndsu.edu wrote:
>>
>>>
>>>
>>> Thanks for ur very useful information.
>>> Can u tell when i am minimizing the sructure and then using that same
>>> minimized structure for pressure, temperature or subsequently force
>>> application should i maintain the same cell basis vectors in every
>>> cases
>>> even if the model size changes at different stages of minimization or
>>> simulation?
>>> debashis
>>>> Hi, Debashis,
>>>>
>>>> You must be new in this list. There are already tons of info
>>>> regarding
>>>> your questions. Here are some quick answers:
>>>>
>>>>> (1) the meaning of periodic boundary condition?
>>>>
>>>> The system of interest is infinitely replicated in x, y, and z
>>>> directions -- for simple box shaped cell. for other complicated cell
>>>> shapes, the replication is more complex.
>>>>
>>>>> (2)How NAMD utilize the periodic boundary condition in it's
>>>>> simulation?
>>>>
>>>> For biomolecules, PBC is not physical. Maybe PBC makes more sense
>>>> for
>>>> crystals. NAMD, or any other MD software, uses PBC as an artificial
>>>> way
>>>> to speed up numerical evaluation of electrostatic forces. Fast
>>>> electrostatics algorithms include MultiPole, MultiGrid, and PME.
>>>> NAMD
>>>> implements PME and Multipole.
>>>>
>>>>> (3)How the value of periodic boundary condition or cell basis
>>>>> vector
>>>>> influence the simulation result.
>>>>
>>>> If you solvate your molecule with a box of water, you better make
>>>> cell
>>>> basis vectors such that they are just a little bit (half angstroms
>>>> or
>>>> 1
>>>> AA) bigger than the water box. If you make cell basis vectors too
>>>> big,
>>>> you introduce artificial vacuum in your system. Your waters will
>>>> fill
>>>> in the bigger box and the density will become smaller. Or small
>>>> bubbles
>>>> will form along some directions, usually close to the boundaries.
>>>>
>>>> Besides PBC, NAMD also supports cylindrical, spherical and vacuum
>>>> boundary conditions. Setting up BCs is dependent on your needs and
>>>> the
>>>> particular systems you work with.
>>>>
>>>> Best,
>>>>
>>>> Marc
>>>>
>>>>> Thanks.
>>>>> Debashis
>>>>>
>>>>>
>>>>>
>>>>> -------------------------------------------------------------------
>>>>> --
>>>>> --
>>>>> DEBASHIS SIKDAR
>>>>> Ph.D. Student
>>>>> Department of Civil Engineering
>>>>> North Dakota State University
>>>>> Fargo, North Dakota-58105.
>>>>> U.S.A.
>>>>> Phone(Office):701-231-6491.
>>>>> Phone(Residence):701-231-4069.
>>>>> E-mail: Debashis.Sikdar_at_ndsu.edu
>>>>> Sikdardebashis_at_rediffmail.com
>>>>> -------------------------------------------------------------------
>>>>> --
>>>>> --
>>>>>
>>>>
>>>>
>>>
>>>
>>> ---------------------------------------------------------------------
>>> --
>>> ***** Keeping aim at low profile is some sort of criminal offence
>>> *****
>>> ---------------------------------------------------------------------
>>> --
>>> DEBASHIS SIKDAR
>>> Ph.D. Student
>>> Department of Civil Engineering
>>> North Dakota State University
>>> Fargo, North Dakota-58105.
>>> U.S.A.
>>> Phone(Office):701-231-6491.
>>> Phone(Residence):701-231-4069.
>>> E-mail: Debashis.Sikdar_at_ndsu.edu
>>> Sikdardebashis_at_rediffmail.com
>>> ---------------------------------------------------------------------
>>> --
>>>
>>
>>
>
>
> -----------------------------------------------------------------------
> ***** Keeping aim at low profile is some sort of criminal offence *****
> -----------------------------------------------------------------------
> DEBASHIS SIKDAR
> Ph.D. Student
> Department of Civil Engineering
> North Dakota State University
> Fargo, North Dakota-58105.
> U.S.A.
> Phone(Office):701-231-6491.
> Phone(Residence):701-231-4069.
> E-mail: Debashis.Sikdar_at_ndsu.edu
> Sikdardebashis_at_rediffmail.com
> -----------------------------------------------------------------------
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:39:14 CST