July 11, 2019 at 9:17 am #30481
First of all thank you so much for the help you providing to the research community in their work.
i have started using BRAPH in my work for FC of EEG data. I have already computed the connectivity matrix (by PLV method) using custom MATLAB scripts. Now i want to do the graph analysis using BRAPH directly from the already computed connectivity matrix. On this forum I have seen the discussion about “how to input the connectivity matrix by Mite” and I follow the steps mentioned. I can successfully load connectivity matrix in the eeg cohort panel. after that i proceed for the graph analysis, but I am confused at that stage. The video tutorial is explained for the eeg time series instead of connectivity matrix, where they define the community structure and calculate. Do I also need to define the community structure after inputting the connectivity matrix?
RabnawazJuly 16, 2019 at 5:48 pm #30823
Thank you for the kind words; I really hope that Braph can prove to be a very useful tool for you.
If you can already load the connectivity matrix, you can proceed to the graph analysis without any additional adjustments. While the videos are indeed explaining the process when the matrices are built from time series, all the steps mentioned in the videos regarding the graph analysis are relevant and analogous for your case also (especially the usage of BUD, BUT and WU graphical user interfaces).
Regarding your specific question, in order for the analysis to work properly you would need to define a community structure. However, if you do not have any structure in mind (defined by any previous knowledge or other considerations) then you could just define a default structure by clicking on the corresponding button. This would basically define a structure for Braph which puts all the nodes in same community initially, but is calculated dynamically. In other words, the default structure tells Braph to recalculate the structure whenever such action is needed (for example, in the calculation of participation coefficient or modularity) via the Louvain algorithm.
I hope this information helps you. However, if you need further assistance, please let me know and I will try to help.