Statistics with GCA result

Submitted by WHJUNG on Mon, 02/21/2011 - 12:38

Dear REST-GCA experts,

I really appreciated your work about REST toolkit.
I have been performing granger causality analysis with resting-state fMRI data in healthy controls and patients using REST-GCA toolkit. I performed both voxel-GCA and ROI-wise GCA.

I am wondering how to conduct statistical analyses, e.g. within-group analysis and between-group comparison with the results acquired using REST-GCA toolkit?
I think that we can not use one-sample t-test or two-sample t-test with the results.
Do you have any description like ppt file to do statistics?

I saw your files, Usage_of_REST-GCA_0.ppt and GCA_readme_New_Version_0.doc, but it didn't describe how to do statistical analysis.

Your reply would be greatly appreciated.

Thank you very much in advance for your help.

Submitted by YAN Chao-Gan on Mon, 02/21/2011 - 22:37

Hi!
For the coefficient-based GCA, you may perform tests on the coefficients directly. See: Hamilton, J.P., Chen, G., Thomason, M.E., Schwartz, M.E., Gotlib, I.H., in press. Investigating neural primacy in Major Depressive Disorder: multivariate Granger causality analysis of resting-state fMRI time-series data. Mol Psychiatry.

For the residual-based GCA, ZANG Zhen-Xiang has a work to convert this value into normal distributed. You can see it from the next release of REST.

Best wishes!

Submitted by WHJUNG on Tue, 02/22/2011 - 08:47

Thank you for your kind comment. It were a big help.
I would use REST-GCA toolkit with our data.
It is cited this software in my manuscript.