Hi Andrew! 

Thanks for all your responses! I just have one last question. What if I want to model a continuous predictor and check whether there exists a linearly increasing subgraph? 

Again I have a repeated measure design across 2 drug conditions but I have ~20 FC matrices for each drug condition per subject. I would have 2 questions:

1. It is often recommended to demean data across ALL participants. I am unsure why though. Wouldn't it make more sense to demean for each subject across drug conditions to account for their individual differences?
   
2. Would a design matrix like this then be correct? 
   

0 -2 1 0 0 0 ....

0 3 1 0 0 0 .... 

1 -0.5 1 0 0 0 ....

0 -8 0 1 0  0 0 

The first column denotes drug (0/ placebo; 1/drug) 
The second column is the per subject demeaned behavioral rating
Column 3 denotes that the first 3 entries belong to participant 1 
Column 4 denotes that the 4th entry belongs to subject 2 
The zeros afterwards would have length(subejct) 

Thank you for your help! 

Cheers, 

Max