No, the cif file (xray i assume) is not suitable for direct calculations. There will likely be steric issues of some kind (e.g. bad C-H bonds or weird angles). Optimize with a cheap, but robust method (e.g dispersion-corrected DFT) and then apply SAPT at the level you need, or can afford.
I would recommend pbeh-3c for the optimization and a SAPT0 calculation to get started.
You might want to check the works by Sure and Grimme. I think they did CB7 in the deltaG blind challenge.