Study-Specific Number of Permutation?

[Jacob Bunyamin]

Dear SnPM experts, 

Happy new year!

I am currently working on single-subject structural VBM analysis comparing with 68 controls. When I ran the SnPM (I picked the Yes (slow, may create huge SnPM_ST.mat. file), we got the following error: 

Failed  'Inference'

Error using snpm_pp (line 834)

Unable to read MAT-file /<workDIR>/SnPM_ST.mat. File might be corrupt.

From what I read in this group, one of the solution is to decrease the number of permutation (we used the default 5000). 

Is there any formula to determine the number of study-specific permutation? Is 5000 too high for our case (is it more suitable for fMRI rather than structural MRI?)

Thank you, 

Kind regards, 

Jacob

[Jacob Bunyamin]

Good afternoon, 

I tried to reduce the permutation count to 1000 but the same error message still appears. Is there anything I need to do to fix this issue?

Thank you, 

Kind regards, 

Jacob