[gmx-users] diffusion coefficient: apparently, g_msd messes up the MSD due to PBC

[David van der Spoel]

On 12/10/11 3:23 AM, Matthew Zwier wrote:
> Hi Ruhollah,
>
> A while ago on the list there was a discussion of extreme memory use
> and possibly-incorrect results from g_msd under some conditions.  The
> problem could be worked around by imaging the trajectory with trjconv
> first to remove jumps across the box, then running g_msd on the
> results.  Perhaps you'll have some more success with that workflow.
>
All of this is not necessary.

You should first try to do more tests before you claim a bug. For 
instance, compute the msd of a single Ca atom.

Furthermore you do not state how your simulation was done and how long 
it was. Since you have a single molecule statistics will be poor unless 
you simulate for 100 ns or so.
> Cheers,
> Matt Z.
>
> On Fri, Dec 9, 2011 at 6:30 PM, Ruhollah Moussavi-Baygi
> <ruhollah.mb at gmail.com>  wrote:
>> David,
>> thank you for your reply.
>>
>> but, g_msd gives a diffusion coefficient which is one order of magnitude
>> higher than expected for a small polypeptide (15 res long with gyration Rg ~
>> 0.8nm) with the same size. this is command line i used:
>> g_msd-mpi -f  md_mine_2nd.xtc -s  md_mine_2nd.tpr -o  msd_2nd.xvg
>>
>> when visualizing, I can see that the polypeptide crosses the boundary
>> several times. apparently,  g_msd counts these 'artificial' jumps in MSD
>> calculation.  and, that's probably why the diffusion coefficient is
>> calculated such high (i.e. 500 um2/s for a small protein).
>>
>> if this is the case, it is a bug in g_msd I believe.
>>
>> is there any way to get round this problem?
>>
>> ---------- Forwarded message ----------
>> From: Anonymous<rossen at kth.se>
>> Date: Fri, Dec 9, 2011 at 12:17 AM
>> Subject: Comment for Forum topic: diffusion coefficient
>> To: ruhollah.mb at gmail.com
>>
>>
>> Hi ruhollahmoussavi-baygi,
>>
>> Comment by davidvanderspoel: diffusion coefficient
>>
>>
>> On 2011-12-09 08:34, Mark Abraham wrote:
>>> On 9/12/2011 5:53 PM, Ruhollah Moussavi-Baygi wrote:
>>>> I want to calculate the diffusion coefficient of a small polypeptide
>>>> with g_msd (see here
>>>> http://www.gromacs.org/Documentation/How-tos/Diffusion_Constant [1])
>>>>
>>>> because of periodic boundary condition, when the peptide goes out of
>>>> the right side it comes in from left side, which leads to an
>>>> artificial displacement. this will give rise to a non-realistic MSD.
>>>>
>>>> is g_msd intelligent enough to avoid this artificial displacement?
>>>>
>>>
>> Yes it is.
>>
>>
>>> Not sure about g_msd, but you can take care of it yourself with a
>>> workflow derived from the information here
>>>
>>> http://www.gromacs.org/Documentation/Terminology/Periodic_Boundary_Condi...
>>> [2]
>>>
>>>
>>> Mark
>>
>> Read more
>> http://support.scalalife.eu/content/diffusion-coefficient#comment-103
>>
>>
>> --
>>
>> This is an automatic message from ScalaLife To manage your subscriptions,
>> browse to http://support.scalalife.eu/user/291/notifications You can
>> unsubscribe at
>> http://support.scalalife.eu/notifications/unsubscribe/sid/131?signature=c1e841f7c725c82291ca5d3191044f3c
>>
>>
>> [1] http://www.gromacs.org/Documentation/How-tos/Diffusion_Constant
>> [2]
>> http://www.gromacs.org/Documentation/Terminology/Periodic_Boundary_Conditions
>>
>>
>>
>>
>> --
>> Best,
>> Ruhollah Moussavi-Baygi
>>
>> --
>> gmx-users mailing list    gmx-users at gromacs.org
>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>> Please search the archive at
>> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
>> Please don't post (un)subscribe requests to the list. Use the
>> www interface or send it to gmx-users-request at gromacs.org.
>> Can't post? Read http://www.gromacs.org/Support/Mailing_Lists


-- 
David.
________________________________________________________________________
David van der Spoel, PhD, Professor of Biology
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596,  	75124 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++