[gmx-users] RIMSP calculation

[R.S.K.Vijayan]

Hi Gromacs users

I am trying to measure the similarity of the subspace sampled by a
mutant protein-Ligand complex versus a wild type complex  based on the
eigen vectors computed using PCA.

 I was wondering, does and handy  scripts does exists to compute the
root mean square inner product ( RMSIP). It would be really thankful
if any one could provide me a  script for RMSIP calculation or kindly
let me know if there exists any way to compute RMSIP directly using
Gromacs.

My second question is ( sorry as this is a general question)

I have come across many papers where they employ RMSIP to quantify the
similarity between PC1 and PC2 derived by essential dynamics and in
turn use the values to prove the convergence of conformational
sampling. Is this a well settled proposition  to drive home the point
that the simulation has converged or should i compute the cosine
content to prove the same. Which of this is a superior metric ?


Thanking in advance

Regards
Vijayan.R