Store the trajectory coordinates as a numpy array, calculate cross-correlation matrix . . .

https://pubmed.ncbi.nlm.nih.gov/1749773/

. . . and visualize it as a heatmap with matplotlib.

Hello Ajit,

Concerning your question:

"Essential Dynamics Analysis is based on the computation of the elements of the positional fluctuations covariance matrix of protein atoms"

The use command gromacs is:

gmx covar –f traj.xtc –s reference.tpr –b start –e end –ascii

You see more details in the following link

https://materials.prace-ri.eu/466/9/MDTrajectoriesHPMDC15.pdf

I find a material that may be of interest in making your question

http://www3.mpibpc.mpg.de/groups/de_groot/compbio/p4/index.html

And other resources complementary to your questions.

http://www.bpc.uni-frankfurt.de/guentert/wiki/images/9/96/180627_AntwortenFragenTutorialMD.pdf

I hope my resource will help you clarify your doubt.

Greetings.

Use python and create a pandas data frame with your trajectory coordinates.

df = pd.DataFrame(data,columns=['A','B','C']) corrMatrix = df.corr() print (corrMatrix)

and try sci-kit learn for PCA

Hi,

First, you should calculate the covariance matrix of your trajectory with "gmx covar"

Then, you can use our Jupyter notebook to convert Gromacs formatted covariance matrix to cross-correlation matrix and heatmap:

https://github.com/aysebercin/gmx_corr

Cheers

You can calculate PCA with prody, which has a user interface in VMD. You can calculate cross-correlation matrix with correlationplus (https://github.com/tekpinar/correlationplus).

thank you Martin Klvana Nicolas Torres Anuraj Nayarisseri A. Berçin Barlas Mustafa Tekpinar