AFNI does not assume the Talairach space or the MNI space or any other space for that matter. If you are using the @auto_tlrc script to transform data from one space to another, the output data will be in the space of whatever template you have used. AFNI's whereami program on the command line and in the AFNI GUI will show the coordinates converted among a set of desired spaces with defaults set to TLRC, MNI and MNI_ANAT. In each case, there should be user control of spaces that allows for any sort of space or atlas, even those for various animal spaces and atlases. See this link for more information:
Several probabilistic atlases are provided with AFNI from the Desai group and the Eickhoff-Zilles group. You may use the probabilistic atlases in the way you have said, but this will be similar to the maximum probabilistic maps that are also provided and generated from the same atlases. A lower threshold of say 10% would take more advantage of the probabilistic atlases. Multiple regions will be indentified using this method. You may also use the larger radius results or the cluster mask overlap results to identify potential regions associated with locations in your subjects.
The FDR results show the expected fraction of false positives at a particular q-value. One would expect many of the false positives to be in small clusters, but that's not guaranteed. The family-wise error correction may be used instead to correct using cluster size (3dClustSim and AlphaSim).
History of AFNI updates
