Hi,
On Tue, Dec 27, 2016 at 9:07 PM, wei <lwstudy.sina.com> wrote:
> While using cpptraj to calculate D, I also want to know whether we should image atoms. According to my test, image and noimage give two totally different results.
When you calculate diffusion you either need to remove imaging (via
'unwrap') or account for it in your calculation. The reason is that
when a molecule is "imaged" it appears to jump across the unit cell in
a single frame, and thus it has an artificially high velocity. As of
version 16 the 'diffusion' command in cpptraj will automatically
account for imaging in both orthorhombic and non-orthorhombic cells
(ptraj and previous versions of cpptraj could only do orthorhombic
cells internally). The 'noimage' keyword should only be used for a
trajectory with no imaging, otherwise the diffusion calculation will
be wrong.
> In Amber16 user manual, it is said that 'In order to correctly calculate diffusion molecules should take continuous paths, so imaging of atoms is autoimatically performed.' This is quite confusing. Because If we wrap the atoms back into box, how can we calculate the real mean square displacement? Because we might underestimate the movement of the molecule.
By default, the cpptraj (version 16) 'diffusion' command internally
keeps track of any 'jumps' and corrects for them.
-Dan
> Any suggestions? Thanks in advance for your help!
> Regards,Liu Wei
>
>
>
> --------------------------------
>
>
> ----- 原始邮件 -----
> 发件人：Huang Jing <jing.huang8911.gmail.com>
> 收件人：lwstudy.sina.com, AMBER Mailing List <amber.ambermd.org>
> 主题：Re: [AMBER] Distance diffusion calculation in ptraj and cpptraj
> 日期：2016年12月27日 21点04分
>
> Hello,I am still a beginner of amber ; the unit of MSD would be Angstrom^2 ;in my opinion, no scaling factor is for mean square displacement and "2n n=3" is needed for the calculation of diffusion coefficient; the calculated diffusion coefficient is the "sug_O[Slope]/2n" n=3,Perhaps my answer is not all correct and you need to examine them,as I am still a beginner,I followed up the official amber tutorial : http://ambermd.org/tutorials/advanced/tutorial15/Tutorial2.xhtml
> jing
> Dr. Jing HuangInstitute of Chemistry and the Lise Meitner-Minerva Center for Computational QuantumChemistry, Hebrew University of Jerusalem, Givat Ram Campus, Jerusalem 91904, Israel
>
>
> On Tue, Dec 27, 2016 at 9:58 AM, wei <lwstudy.sina.com> wrote:
> Dear all,
>
> I want to calculate distance diffusion coefficient D for a sugar molecule in bulk water. I have tried ptraj and cpptraj analysis tools to realize this calculation respectively.
>
> After running the ptraj script, I got five output files (diff.dat_x.xmgr, diff.dat_y.xmgr, diff.dat_z.xmgr, diff.dat_r.xmgr, diff.dat_a.xmgr). Let me display few lines of data deposited in diff.dat_r.xmgr:
>
> 3990.000 1990.5193992.000 2023.4453994.000 1913.1503996.000 1941.7723998.000 1964.5964000.000 1982.162
>
> 1. What does the second column represent and what are their unit? It seems that the second column lists MSD values, but have it been scaled by a factor (for instance, 2n, n=3 in the case of r)?
>
> Let me also paste the calculated D-values by cpptraj:
>
> #Set sug_O[D] sug_O[Slope] sug_O[Intercept] sug_O[Corr] sug_O[Label] 1 0.3546 0.2128 521.8718 0.7209 sug_O_AvgDr 2 0.4605 0.0921 -145.4334 0.6208 sug_O_AvgDx 3 0.3895 0.0779 492.8230 0.5437 sug_O_AvgDy 4 0.2138 0.0428 174.4822 0.5268 sug_O_AvgDz
>
> 2. I was wondering what is the unit of the calculated D values on the second column?
>
> Thanks in advance!
>
> Regards,Liu Wei
>
>
>
>
>
>
>
> --------------------------------
>
>
>
> _______________________________________________
>
> AMBER mailing list
>
> AMBER.ambermd.org
>
> http://lists.ambermd.org/mailman/listinfo/amber
>
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
--
-------------------------
Daniel R. Roe
Laboratory of Computational Biology
National Institutes of Health, NHLBI
5635 Fishers Ln, Rm T900
Rockville MD, 20852
https://www.lobos.nih.gov/lcb
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Dec 28 2016 - 10:00:04 PST
