Hi Lukas,

Regarding your first question, yes, in those results/preprocessing/ROI_Subject*_Condition001.mat files the "data" variable contains the denoised BOLD signal timeseries for each individual ROI, and the associated ROI names are listed in the "names" variable (the raw data is instead stored in the files data/ROI_Subject*_Session001.mat)

And regarding your second question, yes again, those Z matrices contain typically the Fisher transformed correlation coefficients computed from the denoised timeseries (to be precise this will depend on the options that you selected for those 1st-level connectivity analyses, if you kept the default settings choosing bivariate analyses and correlation measures then yes, Z will be Fisher transformed correlation coefficients between each pair of ROIs, but for other options these same Z matrices may contain regression coefficients instead of fisher transformed correlation coefficients, or semipartial correlation or multivariate regression measures)

Hope this helps

Alfonso

Originally posted by Lukas Van Oudenhove:

Hi Alfonso,

I was wondering whether, when I have not checked the "create confound-corrected timeseries" box in the Setup/Options tab of an RRC analysis, CONN still stores .mat files with denoised timeseries for each ROI (I don't need .nii maps)?

More specifically, the results/preprocessing dir in my CONN project contains ROI_Subject001_Condition000.mat and ROI_Subject001_Condition001.mat files, with the latter containing data, names, and d1data and d2data cell arrays with time series for GM, WM, CSF, and all ROIs, but not sure whether these are raw or denoised?

Further, results/firstlevel contains resultsROI_Condition001.mat, with a 3d double called Z, with ROI-to-ROI connectivity matrices for each subjects. Given the name Z, I presume that these are Fisher r-to-z transformed matrices, but can you please confirm? Are these calculated on denoised timeseries?

Finally, would you be willing to look at my remaining issue in an earlier post on denoising settings for gPPI analysis? Remaining problem is that the first level tab does not offer me the gPPI option from the dropdown menu (I performed SBC analysis only since I presume this is all I need for gPPI)?

Thanks a ton in advance for all the help (and the toolbox)!

Best wishes,

Lukas