Hi, Jasmin-
There are a couple different ways to estimate RSFC parameters from resting state data. You *can* do it directly (and validly) with afni_proc.py, *if* you are not including any censoring within the processing. This is because the built-in method is Fourier-transform based, and an assumption for that is to have uniformly sampled data. When no censoring has been used, that would be true, because the FMRI data are acquired with a single TR (=constant sampling rate). When censoring occurs, the data are no longer uniformly sampled, because there are now "gaps" between acquired data, so it is not just "1 TR" of time between each data point to be analyzed.
Most resting state processing includes censoring during processing, because subject motion affects the time series values so steeply.
If you include censoring in your resting state processing, then you would not want to calculate/use RSFC parameters until after the processing were completed. At that point, you could use the pairing of AFNI programs 3dLombScargle and 3dAmpToRSFC to estimate RSFC parameters. *Importantly*, in this route, you would also *not* want to include LFF-bandpassing in your processing, because several RSFC parameters are ratios of LFF magnitudes to unfiltered magnitudes.
--pt
