> Use the same strategy with p-value: for a specific
> q-value, find the corresponding p- or t-threshold,
> and then use 3dcalc, 3dclust, or 3dmerge.
The problem is that the p -> q mapping changes depending on the dataset. In one dataset q=.01 is p=.0123; in another, q=.01 is p=.00018. Is there a way to extract this p->q mapping from a dataset so I can implement your suggestion?