Raphael Fernandes Casseb, PhD University of Campinas, Brazil (Enigma-PD)

We have participated in other multicenter projects and we were aware that projects at the very beginning are not 100% fault-proof. We indeed had a few difficulties in setting up the computing cluster, because our system runs an older Linux version and we did not initially know all the pipeline parameters we would have to set. We reached out to the researchers and developers—especially Eva, Michelle, and Nikhil—who were extremely accessible and supportive throughout the process. We initially had a meeting with Eva who was quite attentive and helpful in pointing directions on how to set up some of the BIDs conversion scripts. Later, we had trouble installing apptainer, and Michelle and Nikhil gave us some hints on what we could do. Finally, we had a few issues with nippopy scripts. Michelle and Nikhil kindly helped us via online meetings and followed up on discord to guarantee everything was fine. In terms of workload, I initially expected the setup to take a little time. In reality, it took more time than planned, but their team provided all the assistance we needed. The outcome has been very positive and we were able to run all the analyses we wanted. Regarding workshops and tutorials, I think they could keep doing exactly as they are doing: brief meetings with initial orientation and let users try it for themselves. Then, they can work with users individually to sort out specific problems. From my perspective, there is no point in hosting long sessions, considering that each user may be running a different system.