[gmx-users] diffusion, g_msd and trestart

[David van der Spoel]

Enemark Soeren wrote:
> 
> -----Original Message-----
> From: gmx-users-bounces at gromacs.org
> [mailto:gmx-users-bounces at gromacs.org] On Behalf Of David van der Spoel
> Sent: Monday, July 27, 2009 5:12 PM
> To: Discussion list for GROMACS users
> Subject: Re: [gmx-users] diffusion, g_msd and trestart
> 
> Enemark Soeren wrote:
>> Dear all,
>>
>> I am trying to calculate diffusion coefficients for a system of (30) 
>> glycine molecules in a box with (1000) water.
>>
>>  
>>
>> After running:
>>
>> g_msd -f md.xtc -s md.tpr -mol -n molindex.ndx
>>
>>  
>>
>> I get a list of diffusion coefficients - one for each glycine
> molecule. 
>> I then calculate the average and the stdev:
>>
>>  
>>
>> Average=1.004
>>
>> Stdev=0.7661
>>
>>  
>>
>> My problem is that the stdev is very great! More than +/- 75% of the 
>> average value.
>>
>> Why is that?
> 
> Short simulation most likely. Since these things are big it takes a long
> 
> time to equilibrate.
> 
> My simulation is 10ns long with xtc saved every 1 ps. I thought I had
> enough data!?
> How long is "long time"?
> 
>>  
>>
>> Also, what is the reason for the trestart=10 default value? Shouldn't
> it 
>> be better to have trestart=1, if my xtc file is saved every 1 ps?
> 
> No, because there is correlation in the system. Ideally you would take 
> this time to be longer than the correlation time. You could try to find 
> an estimate of that from the velocity autocorrelation.
> 
> So, I must run velocity autocorrelation to find out when the velocities
> become uncorrelated, and then use this time as my trestart?
> .

That would be a way (and the VACF also gives you a diffusion constant), 
you can also check the histogram visually for different run times. It 
seems that your solution does have a lot of amino acids, which probably 
means high viscosity and slow diffusion, and hence long time to 
equilibrate. Looking at the system MSD for all the Gly may also show you 
when the MSD curve becomes linear, you can use this as well as an 
estimate for the trestart.


> Thanks,
> Soren
> 
>>  
>>
>> Thanks,
>>
>> Soren Enemark
>>
>>
>>
> ------------------------------------------------------------------------
>> _______________________________________________
>> 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/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/mailing_lists/users.php
> 
> 


-- 
David van der Spoel, Ph.D., Professor of Biology
Molec. Biophys. group, Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205. Fax: +4618511755.
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se