[gmx-users] Principale component analysis , experiment and MD comparison

[Justin Lemkul]

On 6/29/17 2:32 PM, Nawel Mele wrote:
> Hi all,
> 
> I would like to compare different MD trajectories with an ensemble of NMR
> structures by PCA. I performed the PCA by projecting all my MD structures
> and NMR ensemble onto the eigenvectors of my NMR ensemble then plot the
> result.

This method is correct, because you're comparing the structures in terms of the 
same PCs.

> However, I also performed PCA for each MD trajectories separately and NMR
> separately and plot the result in one figure to see the difference between
> the two methods and now I am confused on the process to compare different
> MD with experimental data or even with a random generator.
> 

This method is incorrect because you may be generating different PCs between the 
MD ensemble and the NMR ensemble.

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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