Hi John,
You can use the FEAT gui to configure and run your preprocssing requirements (temporal filtering, spatial smoothing, registration, etc). When FEAT completes, the preprocessed 4D FMRI data, in the subject's native space, can be found in your .feat directory, and is calld 'filtered_func_data.nii.gz'.
You can transform this file to MNI152 space with the applywarp command (assuming that you have used non-linear registration):
applywarp -i filtered_func_data.nii.gz -r $FSLDIR/data/standard/MNI152_T1_2mm.nii.gz -o ffds.nii.gz -w reg/example_func2standard_warp.nii.gz
Cheers,
Paul