I ran a connectome script, which generated streamlines between ROIs from the brainnetome atlas, and then executed a script that derive controllability from it. The outcomes from the script are a column per participant and a controllability value for each of the 246 ROIs. I have this before and after administration of a drug. ( I know controllability values are not that common but I guess it would be similar to a matrix of connectivity values) . How can I enter these values (csv file) into a model that would test controllability differences before and after the drug? Which I assume woudl be a basic t-test
To clarify, does each participant have two controllability values per ROI (one pre-treatment and one post-treatment), resulting in 492 values per individual across the 246 ROIs? What research question(s) are you trying to address with the data?
sorry I thought I answered but looks like it did not go through. I have one controllability value per ROI and 6 participants at pre and post intervention . So my csv file is 256*12. I just want to test whether the controllability values changed after the intervention in any of those ROI.
You have two main modeling options depending on how you wish to handle the multi-region structure:
Conventional paired t-test: You can perform a paired t-test separately at each ROI. Since this method treats each region in isolation with no shared information, you need to control for multiple testing across all ROIs using an adjustment method like False Discovery Rate (FDR).
Hierarchical model via RBA: Alternatively, you can fit a single model across all ROIs simultaneously using the AFNI program RBA, entering the contrast of the two controllability values per participant as input. Because RBA incorporates population-level structure and partial pooling across regions, no post-hoc adjustment for multiplicity is required.
The
National Institute of Mental Health (NIMH) is part of the National Institutes of
Health (NIH), a component of the U.S. Department of Health and Human
Services.