[gmx-users] Dynamic Cross Correlation Map or Cross-correlation matrix

atanu das samrucu at yahoo.co.in
Sat Sep 10 05:15:04 CEST 2016


Hi All,
I have been trying to evaluate the dynamical cross correlation matrix or commonly known as cross-correlation matrix from a trajectory. The g_covar program gives the cross-covariance matrix as covara.xpm. However, that's not what I want to estimate. I want to calculate the C-alpha based cross-correlation matrix where the correlation values vary from -1 (anti-correlated) to +1 (correlated). There are few open gromacs-compatible programs available e.g. g_covar (gromacs website, additional programs), g_correlation (max planck website); however none of them works properly (version issues, OS issues).
There is also a suggestion on the web to take the covar.dat ascii file and then post-process it. However, in that case it would be helpful to have a covara.dat from covara.xpm as covara.xpm differs from covar.xpm (covara.xpm is more advantageous in case of cross-correlation matrix estimation) as in the former case one gets the values foreach atom pair as the sum of the xx, yy and zz covariances rather than de-convoluted in 3N terms. 
Please advice. 
Thanks a lot in advance.
Atanu  


More information about the gromacs.org_gmx-users mailing list