How to construct cij matrix with non-linear mutual information approach

[Former user]

Dear sir,

In your recent journal paper "Dynamic coupling and allosteric networks in the alpha subunit of heterotrimeric G proteins" , you have mentioned that your crosscorrelation calculation is based on recently developed non-linear mutual information approach .As far i know, the cna-vintage tutorial(protein structure networks with Bio3D) which i am following for the calculations deals with pearson-like approach. so how can i use the non-linear mutual information approach in my calculations?. More specifically, what will be the code for that?

thanks

[Xinqiu Yao]

Hi,

We used the linear Pearson correlation and it has be clearly described in the paper (See the "Experimental Procedures" section). Currently, we don't have the function in Bio3D (but will likely be implemented in future) for non-linear mutual information calculation. A function for 'linear' mutual information calculation called lmi() does exist and you are encourage to give it a try.

[Barry Grant]

this is not a bug, rather a Q&A