There are a fair number of details that are needed to know exactly what you are doing, so a whole script might be helpful to see. Overall, I think it looks okay, but it's hard to say without the actual commands. Here is a script for doing something similar. It will probably be needed to be adjusted for different grids. Consider using the -master_.... options in the alignment procedures, so that interpolation only happens once and onto the grid you want.

[afni.nimh.nih.gov]

In the FCON-1000 data, there was a similar problem of left-right flipping between anatomical and EPI data. This kind of problem can often be caused by using Analyze format intermediate files or incorrect NIFTI files. See this posting regarding how to use align_epi_anat.py as a check for this.

[www.nitrc.org]

Sorry, our upload data link doesn't work yet. If you would like, you can send data to me using other on-line storage.