Dear Alfonso,
I used CONN to analyze a set of resting-state and task-induced data in two groups of participants, and found significant negative correlation in both groups. One of reviewers suspects that the small z scores (0< - 0.1) do not indicate an anticorrelation but might possible indicate a small phase shift of the low-frequency periodic signal. I want to ask whether CONN-fMRI has considered this question? 
Best,
Ivan