[gmx-users] error in umbralla sampling step 6

[Justin Lemkul]

On 10/27/13 11:23 AM, sunyeping wrote:
>
>
> Yeping Sun
> Institute of Microbiology, Chinese Academy of Sciences
> ------------------------------------------------------------------
> 发件人：Justin Lemkul <jalemkul at vt.edu>
> 发送时间：2013年10月27日(星期日) 20:27
> 收件人：gromacs <gmx-users at gromacs.org>
> 主　题：Re: 答复: [gmx-users] error in umbralla sampling step 6
>
>
> Please keep the discussion on the list.
>
> On 10/27/13 5:09 AM, sunyeping wrote:
>> Dear professor Lemkul,
>>
>> For warning 1:
>> I use the npt_umbrella.mdp file you provided in step 6 in the umbralla sampling
>> tutorial. You told us to start by running a brief NPT equilibration in each
>> window using this mdp file. If Parrinello-Rahman pressure coupling should be
>> used, why did you put it in the mdp file?
>>
>
> I know the tutorial system is robust enough that it won't matter.  That is not
> necessarily true for whatever other system you come up with.  Be careful when
> extrapolating something that works in a tutorial with something that is
> untested.  The general logic is the same (you need to equilibrate for a little
> while first), but you have to think a bit about the proper settings.  There is
> also no guarantee that the length of time I set in the tutorial simulation
> processes (equilibration, SMD, and umbrella sampling) will necessarily be
> sufficient for whatever you're doing.
>
>> For warning 2:
>> I check the gro files using vmd but I cann't find any molecule broken? How can I
>> find the brroken molecule?
>
> The only way to find them is to look.  I have no special advice for this.
>
>> I add -pbc whole option to the  trjconv as following:
>>
>> trjconv -s pull.tpr -f traj.xtc -o conf.gro -sep -pbc whole
>>
>> then It seems that the problemis solved. Is that a correct wayyou mean to solve the problem?
>>
>
> Then clearly something was broken across PBC, which means there is no real problem.
>
> -Justin

> In case that something was broken across PBC, why there is no real problem? Do you mean molecule breaking will not affect the simulation result?
>

We use the term "broken" very loosely.  The molecule is still intact from a 
topological standpoint.  It is only "broken" in the way that we see it.  The 
dynamics of a given molecule are the same if the molecule is "intact" in the 
"center" of the box (also a misnomer in a periodic system) or "broken" and a 
boundary.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441

==================================================