Averaged connectome across subjects

[Alice Pierini]

Hello everyone, 

I have read in some papers and conversations in this forum that sometimes averaging the SC matrixes of different subjects boosts the results regarding simFC-emFC correlation. 

I was wondering what is the most appropriate methodology to do this. 

So far, to extract the SC, I have performed tractography by staying in the subject-specific DWI space, as suggested in the MRtrix documentation. Therefore, tract length matrixes and ROIs centres coordinates are subject-specific as well. 

Is it therefore necessary to perform tractography after the registration of the subjects' DWI volumes to the MNI space? (Which, however, seems to be discouraged in forums such as MRtrix because it invalidates the meaning of streamline analysis and requires additional steps regarding the warping of tracts obtained from tractography analysis).

What do you recommend in this case? Would it be an option, for example, to simply compute the average of the subject-specific tract length matrices?

Thank you very much for your help,
Alice