Combining aCompCor with global signal regression in xcpengine

Dear community,
I have some BOLD data on which I’d like to run connectivity analysis. I preprocessed te data using fMRIprep, and now I’d like to use xcpengine for nuisance regression.

I have decided to use the fc-acmopcor.dsn pipeline. Following Murphy et al. (2020) “Multimodal network dynamics underpinning working memory”, I would like to add Global Signal Regression on top of the aCompCor pipeline.

The way I chose to do so is to edit the fa-acmopcor.dsn file, replacing:

The reason I am worried is that looking at the “confound2” output directory, I see that a file names “sub-011_task_rest_run-1_gsr.1D” was not created.

I would appreciate comments on this approach. Is this the correct approach? Would that regress the global signal in an apropriate manner?
Any advice would be greatly appreciated.


Just pinging this in case anyone encountered a similar problem and has some useful insight.
Many thanks!