I decided to go with using align_epi_anat with lpa cost function to align the template functional image (sss_template) with my subject's functional data (201.breathhold.BOLD.nophymotcorr.tcat.nii). Both were functional images that were not skull-stripped.

align_epi_anat.py -dset1to2 -dset1 sss_template.nii.gz -dset2 201.breathhold.BOLD.nophymotcorr.tcat.nii -dset1_strip None -dset2_strip None -cost lpa

I then used the resulting transformation matrix (sss_template_al_mat.aff12.1D) with 3dAllineate to align the SSS ROI (sss_mask) with my subject's functional data.

3dAllineate -cubic -1Dmatrix_apply sss_template_al_mat.aff12.1D -prefix sss_mask_al sss_mask.nii.gz

Does this approach make sense? How do I know if the alignment was done correctly? I get the following error message if I use the -AddEdge option:

#Script is running:
cd ./; \rm -f AddEdge/*
#++ resampling dset2 to match dset1 data
#Script is running (command trimmed):
3dresample -master ./__tt_sss_template+orig -prefix ./__tt_201.breathhold.BOLD.nophymotcorr.tcat.nii_rs_in -inset ./__tt_201.breathhold.BOLD.nophymotcorr.tcat_ts+orig'' -rmode Cu
** ERROR: Could not resample

Traceback (most recent call last):
File "/Users/Alain/AFNI_verX/align_epi_anat.py", line 3258, in <module>
listlog = listlog_a2e)
File "/Users/Alain/AFNI_verX/align_epi_anat.py", line 2688, in add_edges
d1AE = afni_name("%s%s" % (d1name,d1.view))
AttributeError: 'NoneType' object has no attribute 'view'